Journal of Cleaner Production 237 (2019) 117769

Contents lists available at ScienceDirect

Journal of Cleaner Production

journal homepage: www.elsevier.com/locate/jclepro

Corporate sustainability on causal financial efficiency model in a

hierarchical structure under uncertainties
Shulin Lan a, Chen Yang b, Ming-Lang Tseng c, d, *
a
School of Economics and Management, University of Chinese Academy of Sciences, Beijing, PR China
b
School of Computer Science and Technology, Beijing Institute of Technology, Beijing, PR China
c
Institute of Innovation and Circular Economy, Asia University, Taiwan
d
Department of Medical Research, China Medical University Hospital, China Medical University, Taichung, Taiwan

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

Article history: This study contributes to (1) the financial report attributes through the measurement-based algorithm
Received 19 April 2019 and uses the financial data of the listed firm to test the validity and reliability of the proposed corporate
Received in revised form sustainability model; (2) the impulse response relationship of the four first-level indicators and provides
21 July 2019
a basis for decision-making to improve the efficiency of a firms' financial operations; and (3) this study
Accepted 23 July 2019
Available online 26 July 2019
applies the primary and secondary data from those listed firms. Still, the financial report is the important
sources for the public to obtain operations and accounting information. Few studies are presented
Handling editor. Prof. Jiri Jaromir Kleme
s financial report with hierarchical and causal interrelationship model in considering the corporate
financial sustainability in expert preferences and firm's secondary data together. This study is presented
Keywords: the four parts financial reporting are structured as 1. Earning ability; 2. Efficiency; 3. Debt; and 4. Asset
Corporate sustainability management. The authenticity and reliability of financial reports are inevitably affect the quality of ac-
Financial efficiency counting information provided and establishes a decision-making index system that affects the quality of
Decision making trial and evaluation the financial efficiency according to the four aspects. This study proposes to apply the decision making
laboratory
trial and evaluation laboratory and analytical network process together to handle the hierarchical and
Analytical network process
causal interrelationship. In addition, prior studies are lacking to address in built on the score data of
Financial sustainability
Entropy weights different influencing indicators using panel data vector auto regression model. The result shows the asset
management capability has a positively impact on the firms' profitability and asset-liability.
© 2019 Elsevier Ltd. All rights reserved.

1. Introduction guarantee for the sustainability of a firm. Strategies on financial

sustainability of a firm run through its operation, from financing
As the core of corporate strategy, financial sustainability strat- and investment to the allocation of resources (Wang et al., 2019a,b).
egy is the driving force for a firm's development. Most firms' Sound strategies on financial sustainability benefit firms to better
development track at home and abroad shows that the financial organize different parts from the upper stream to the downstream
strategy of a firm and its sustainability are closely linked. The and make creative plans focused on big pictures and adapt to
change of one factor inevitably influence the other. On the one changes of the market.
hand, the financial support for strategy provides the basis for its The concept of sustainability has been adopted by firms and
sustainability. Financial sustainability helps a firm to properly use used in financial affairs with social development in recent years.
its financial resources and financial information to maintain a sta- Sustainability of corporate finance means the maximization of
ble flow of funds. Through sound strategies on financial sustain- corporate value with highest growth rate based on a stable financial
ability, a firm properly allocates the resources to avoid financial condition without running out its financial resources. Stability is
risks and operational risks. A sustained financial strategy provides a mainly reflected in the rapid and stable cash flow, fast inventory
turnover and low level of accounts receivable. The notion of
financial efficiency is generated in this context. Financial efficiency
* Corresponding author. Institute of Innovation and Circuar Economy, Asia Uni- refers to the best performance of a firms' financial affairs through
versity, Taiwan. investment in a period of time under certain technological condi-
E-mail addresses: lanshulin@ucas.edu.cn (S. Lan), yangchen666@bit.edu.cn
(C. Yang), tsengminglang@asia.edu.tw (M.-L. Tseng).
tions. Financial efficiency can not only show the firms' ability to

https://doi.org/10.1016/j.jclepro.2019.117769
0959-6526/© 2019 Elsevier Ltd. All rights reserved.
2 S. Lan et al. / Journal of Cleaner Production 237 (2019) 117769

allocate resources and pay back its debt, but also its ability in proves the effectiveness of the DEMATEL-ANP method; (3) the
earning profits and its future potential. It is a comprehensive in- entropy weight method is used to objectively study the relationship
dicator of a firms' financial condition (Taqi et al., 2018a; Tseng et al., of these indicators and the impact of them on the efficiency of
2019a; Tseng et al., 2019b). Mostly, the higher the financial effi- financial operation based on the actual operational data of the
ciency of a firm, the more sustained its financial condition is and different indicators gathered in public firms from 2008 to 2017. The
vise versa (Wu et al., 2017; Zhang,2015; Yang et al., 2018). The 41 firms' operating data were selected, and the entropy method
assessment of a firms' financial efficiency is conducive to an was used to identify the importance of each indicator In order to
objective evaluation of the firms' performance and its health objectively measure the impact of these indicators on promoting
development. The first way for firms to develop a firm is to improve the financial sustainability.
the financial efficiency, which contributes to the financing of a firm. Finally, the Panel Data Vector Auto regression (PVAR) model is
This layes a solid foundation in terms of firm's capital to expand its built on the score data of different influencing indicators, and the
production scale. A larger share of market can be obtained and the causal relationship of the primary indicators is tested by impulse
firm maintains its sustainability while being more competitive response analysis to reflect the dynamic impact of these indicators.
(Cheng et al., 2013; Chen et al., 2014; Tseng et al., 2016). Hence, this The model can find the relations among first-tier indicators of
study focuses on four financial ratio indicators as earning ability, financial reporting system. The model analyzes how these factors
Efficiency, debt management, and asset management. influence a firms’ financial management and provides theoretical
Enron and other malignant economic cases showed that the support. The model itself can serve as a starting point for firms to
quality of corporate financial reporting is not only related to the control financial risks. This study identifies the key factors
internal accounting process of a firm, but also with a variety of improving financial efficiency and provides a scientific basis for
external factors. Enron's case taught us a control of financial risks promoting the financial sustainability, strengthening the financial
are necessary to maintain the sustainability of a firm. For example, reporting process control, and improving the quality of financial
accounting principle and external auditing can exert the direct and operation.
indirect influence on the quality of a firms' financial report (Bratten
et al., 2019). The financial risks of firms are difficult to measure due 2. Literature review
to the uncertainty of internal and external factors in the financial
activities of firms. The financial risks of firms run through the whole Financial sustainability Martínez-Ferrero and Frías-Aceituno
process of financial activities, and they influence the income and et al. (2016) required firms to optimize their financial activities
expenditure of a firm. It might also lead to insolvency. The and properly handle various financial relationships under the
improvement of a firms' earning ability, efficiency, debt manage- constraints of existing scarce resources while keeping its creativity.
ment, and asset management can reduce the financial risks of a In this way, a firm can gain the ability to operate in a financially
firm. This shows higher financial efficiency means lower level of sustainable way, gain more profits, promote the sustainable growth
risks. The better one firm performs, the lower the risk of default and of corporate value. The concept of financial sustainability has
bankruptcy. The more financial efficient a firm is, the stronger it received wide attention. Checherita-Westphal et al. (2014)
will be in managing the debt, operating, and earning profits. The emphasized the importance of debt-related fiscal principles and
lower the probability that the actual income of the enterprise de- achieving financial sustainability by maximizing the public debt
viates from the expected return, the lower the financial risk of the ratio of growth through a simple theoretical model. Afonso and
enterprise (Kirschenmann et al., 2016). Jalles (2016) assessed the sustainability of public finances in
Therefore, it is necessary for scholars to improve their methods OECD countries using unit root and cointegration analysis con-
to research problems from both the internal and the external per- trolling endogenous breaks. They control endogenous breaks and
spectives, and analyze the quality of a financial report from more find it is hard to measure a firms' financial sustainability accurately.
broaden perspective. The information distortion and the low Navarro-Galera et al. (2016) think that the income statement is a
quality of financial report be comprehensively treated. Against this good way to assess financial sustainability because it reveals in-
background, the concept of ““financial sustainability” gradually formation about the three dimensions which are revenues, debt
become an important subject in the accounting research and ac- and services proposed by the International Federation of Accoun-
counting practice (Wang et al., 2019a,b). The financial analysis in- tants. Coad and Rao (2008) believed that whether the listed firms’
dex system is established from the perspective of financial innovation strategy is rational can directly determine their financial
management. The financial indicators such as solvency, operational sustainability. A reasonable innovation strategy is beneficial for
capability and development capability of the enterprise are listed firms to gain control of products and effectively increase their
analyzed to monitor the risk signals, so as to propose a feasible risk market share but it can have an adverse effect on short-term
management strategy and reduce the harm. financial performance. Generally speaking, the existing literatures
The contribution of this study are (1) an establishment of eval- are rather limited about studying financial sustainability, and lack
uation measurement systems is based on decision making trial and an index system for multi-dimensional evaluation of corporate
evaluation laboratory and analytical network process (DEMATEL- financial sustainability.
ANP) to handle the hierarchial structure and interrelationships Financial efficiency is an important indicator to measure
among the proposed factors. Debt management is the first among corporate financial sustainability and corporate sustainability
these four factors to influence the quality of financial report supply (Grewatsch and Kleindienst, 2017; Tseng et al., 2019b). However,
chain, followed by efficiency, earing ability and asset management. blindly pursuing financial efficiency can also reduce actual business
The study builds evaluation system that affect the financial efficiency (Edmans et al., 2016). Hence, how to measure the
reporting supply chain optimization from different aspects; (2) financial efficiency of firms and maintain the financial sustain-
with the ANP network structure as its basis, combined with the ability of firms has become a hot topic for practitioners and aca-
expert scoring data, the order of weights of the primary indicators demicians. The existing study relies on financial report to analyze
follows DEMATEL's causal analysis's order of influencing the quality the financial efficiency of listed firms, and uses financial indicators
of the financial reporting supply chain (Tseng et al., 2019b). The such as solvency, operational capability, profitability and growth
order of weights of the indicators obtained from entropy method is capabilities to measure a firms' financial efficiency (Jha, 2011; Tseng
consistent with the results from ANP analysis. The consistency et al., 2018b). Some scholars also focus on the impact of non-
 S. Lan et al. / Journal of Cleaner Production 237 (2019) 117769 3

financial indicators on financial efficiency. For example, Curtis et al. determine the key factors for successful project management, and
(2015) think that corporate financial information has the largest used ANP to study the role of each indicator that affects the final
impact on asset turnover ratio indicators in accordance with US results of the project. DEMATEL-ANP has also been applied in
General Accounting Acceptable Principles Standards' adjustments. customer relationship management partner evaluation
In other words, adjustments of accounting standards affect firms’ (Büyüko € zkan et al., 2017; Tseng et al., 2019b), renewable energy
financial efficiency. resources selection, cleaner production (Govindan et al., 2016), etc.
Financial risk refers to the risk of bankruptcy due to excessive And it has been widely recognized that this model is useful.
financing and poor management which would lead to a firms' Until now, little works, especially that uses quantitative
inability to repay its debts (Cantele and Zardini, 2018). Financial methods, has been conducted on how to strengthen and optimize
risks are closely related to a firms' debt management ability, prof- the financial reporting supply chain. This study tries to firstly
itability, financial operation efficiency, and asset management ca- identify the key factors and their weights to evaluate the financial
pabilities. The level of efficiency determines the financial risk to a reporting supply chain and then provides guidelines for efficient
certain extent. Corporate debt capacity refers to the ability of firms improvement. The financial reporting supply chain is a term used to
to repay their debts on time. Reasonable financial leverage can depict the myriad of players involved in the earning ability, effi-
maintain the sustainability of firms. Excessive financial leverage ciency, debt management and asset management. With many
increase the financial risks of firms (Martellini et al., 2017; Tseng influencing factors in the financial operation, an index system
et al., 2019b). However, due to a variety of factors, firms with low would inevitably include key factors that have mutual relationship
leverage and high profits may also have higher financial risks and affect each other. So proper methods are needed to evaluate the
(Giordani et al., 2011). The profitability refers to the ability of a firm indicators’ effects on the quality of the financial efficiency and
to make a profit in its daily operations (Benitez et al., 2018). This is a financial sustainability.
strong guarantee for a firm to pay its principal and interest on time.
The stronger the profitability of a firm and the larger share of 3. Framework for financial efficiency evaluation
revenue have to pay for the loan principal and interest, so the
financial risk of the firm is negatively correlated with the profit- 3.1. Evaluation model based on DEMATEL-ANP
ability of the firm. Financial operational efficiency indicator is
mainly used to measure the efficiency of the firms' asset manage- There are many factors that directly contribute to the quality of
ment. The efficiency can be measured by commodity turnover rate the above aspects. Those factors should be first found to build an
and crossover ratio (Martínez-Campillo et al., 2018). The more evaluation index system (Fig. 1). This study identified that satis-
efficient a financial operation is, the more profit a firm can make in factory solutions are still needed in four aspects of the financial
a certain period of time (Nguyen et al., 2016), which can improve efficiency: Earning ability, Efficiency, Debt management and Asset
the profitability of the enterprise and reduce the financial risk. management. This authoritative clue can be used as the starting
Asset management capability is the operational capability of one's point to identify the key factors.
business (Ga ^rleanu and Pedersen, 2018). The stronger the opera- This study found that these factors would likely exert different
tional capability is, the more profit a firm can generate from the degree of influence on each other during the investigation of the
investment on certain resources or funds (Hirunyawipada and potential factors. In other words, the factors affecting the financial
Xiong, 2018). In other words, the profitability of a firm also in- efficiency may not be totally independent. This study needs to
crease, and the timely payment of principal and interest also be a involve this characteristics, when evaluating the importance of the
strong guarantee. The financial risks of a firm will be effectively factors. Therefore, in the process of constructing the optimization
controlled. Financial efficiency has a close relationship with index evaluation model of financial efficiency, this study considers
financial risks. The improvement of corporate profitability, financial the logical relationship between the indicators, and combines the
operation capability, debt management capability and asset man- DEMATEL and ANP methods that emphasize the relevance of the
agement capability, which is also known as the improvement of indicators and use the DEMATEL method to effectively sort out the
financial efficiency, is conducive to reducing corporate financial complex relationship between the indicators (Tseng et al., 2019a;
risks. In addition, financial efficiency is also affected by a corporate b). To determine the relationship and intensity of the primary in-
social irresponsibility (Ko €lbel et al., 2017; Tseng et al., 2018b), dicators, then use the ANP method to calculate the relevant weights
macro-economic policy (Fouejieu et al., 2019) and technological of the secondary indicators, so as to enhance the objectivity and
environment (Chae et al. 20 18). preciseness of the evaluation results.
Comprehensive evaluation method, stochastic frontier analysis The ANP method can still provide more objective and precise
method and data envelopment analysis have also been widely used evaluation results in the case of non-independent relationships
to measure efficiency. Analytic hierarchy process is used to the between indicators. The premise of applying the ANP method is to
evaluate the independent hierarchical structure, but it is not suit- scientifically and reasonably clarify the specific logical relationship
able to calculate the weight of indicators that are not independent between the indicators, and the DEMATEL method can provide a
of each other. In order to solve this drawback of the analytic hier- scientific basis for the determination of these relationships. This
archy process method, Lam and Lai (2015a,b) uses the graph theory study establishes a fuzzy ANP network structure, and calculates the
and the matrix tool to analyze the index, calculate the centrality limit super-matrix by comparing the judgment matrix to determine
and cause of each index through the logical relationship between the weight of each indicator in the fuzzy ANP network based on the
the indicators and the direct influence matrix. Gabus and Fontela logical relationship between the indicators obtained by the
(1972) build the DEMATEL causal map to determine the logical DEMATEL causal map.
relationship of indicators and its importance in the system.
DEMATEL and ANP have been widely used in the establishment of 3.2. Building the relationship among financial efficiency evaluation
indicator systems and the identification of key factors. Based on indicators based on DEMATEL
ANP- quality function deployment which combines ANP with
quality function deployment, Lam and Lai (2015a,b) established an This study takes the four direct factors affecting financial report
index system for shipping firms to achieve environmental sus- quality. The evaluation index system consists of two levels: the
tainability. Mavi and Standing (2018) used DEMATEL method to primary indicator (four aspects) and the secondary indicator (the
4 S. Lan et al. / Journal of Cleaner Production 237 (2019) 117769

Building inter-indicator Quantitative indicator

Selected evaluation indicators
relationships based on influence based on ANP
Demtel
Establish a direct- Draw the network
relation matrix structure of ANP
Building relationships between Standardized direct- Impact relationship
indicators between indicators Structure judgment
relation matrix
matrix
The comprehensive
influence relation
Quantify the impact of the Calculated
matrix is calculated
indicator weighting super-
matrix
Indicator analysis

Calculated limit
Get the weight and order of the Indicator analysis super-matrix
indicators

Fig. 1. Financial efficiency evaluation model.

indicator that directly affects the primary indicator). If it is negative, it means that the element is greatly affected
by other factors, which is called the cause factor.
(1) Evaluating the direct influencing relationship between the
primary indicators and constructing a directed graph. If in- T ¼ ½tijn  n i; j2f1; 2; 3; :::::; ng (4)
dicator Ui exerts a direct impact on the Uj , draw an arrow
from Ui to Uj and use the numbers 0, 1, 2, 3, 4 to indicate the X
n

strength of the relationship between the indicators on the D¼ tij (5)

j¼1
ray. Among them, 0e4 represents the influence degree of no
effect, low, medium, high and extremely high.
(2) The normalized direct relation matrix X is calculated by X
n
R¼ tij (6)
Equations (1) and (2).
i¼1

X ¼K  A (1)
(5) Based on D þ R, D-R the mapping of the primary indicator
! causality, the relationship between the primary indicators
1 1 and its importance can be clearly identified.
k ¼ min Pn  ; Pn   (2)
max a
j¼1 ij max i¼1 aij
1in 1jn

3.3. Quantify the influence of financial efficiency indicators based

on fuzzy ANP
(3) Based on the standardized direct relation matrixX, the
comprehensive influence matrix T is obtained by using
(1) According to the causal map of the primary indicators of the
Equation (3).
financial effciency obtained by DEMATEL, combined with the
actual situation of the secondary indicators, the ANP network
X
∞
structure is drawn. The ANP network structure consists of a
T ¼ X 1 þ X 2 þ X 3 þ :::: ¼ X i ¼ Xð1  XÞ1 (3)
i¼1
control layer and a network layer. The control layer generally
includes a target layer and a criterion layer. The network
layer consists of a set of elements (primary indicators) and
(4) Indicator analysis. Investigate the indicator tij of T, calculate elements (secondary indicators) that are dominated by the
the effect degree D, affected degree R and center degree control layer. Peer indicators are independent of each other
D þ R and cause degree D-R of each indicator. Among them, and dominate each other.
tij indicates the degree of comprehensive impact from in- (2) Construct a judgment matrix. It is assumed that the criteria
dicators j to indicator i. The D is the sum of rows in the of the target G in the control layer are G1、G2、 … 、GM, and
comprehensive impact matrix T, indicating the combined the elements in the network layer are C1、C2、 … 、CN,
impact of the indicator on all other indicators. R is the sum of where Ci has elements ei1,ei2, …eini (i ¼ 1,2, …,N), there are
the columns in the comprehensive impact matrix T, indi- elements ejl (j ¼ 1,2, …,N; l ¼ 1,2, …,nj)in Cj. Taking GS(S ¼ 1,2,
cating that the indicator is affected by all other indicators. …,M)in the control layer as the criterion and the element ejl
The sum of D and R of each indicator is called the centrality of in the sub-criteria as the sub-criteria, the other elements in
the indicator, indicating the position of the indicator in the the element set Ci are compared in pairs with respect to the
system and its role. The difference between D and R of each importance of ejl using constructing The pairwise compari-
indicator is called the cause degree of the indicator. If D-R is a son matrix of elements in Ci relative to ejl under the criterion
positive number, it means that the indicator has a greater GS and the normalized feature vector wi1ðjlÞ ;
influence on other indicators, which is called the cause factor. wi2ðjlÞ ::::::winiðjnjÞ , that is, the network element sorting
 S. Lan et al. / Journal of Cleaner Production 237 (2019) 117769 5

vector. In order to compare the relative importance in the 4.2. Efficiency

pairwise comparison, this study uses the 1e9 scale are used
in this study (Saaty and engineering 2004). (Table 1). Simi- The efficiency ratio is mainly used to measure the efficiency of
larly, a sort vector with respect to other elements can be the firm in asset management, that is, to analyze the effective uti-
obtained, resulting in the following matrixWij. Among them, lization of enterprise resources. Efficiency indicators often reveal
the column vector of Wij is the element ei1,ei2, …,eini influ- some of the problems in the operation of firm assets. Efficiency
ence degree on elements in Cj. refers to the analysis of operation efficiency and turnover of the
(3) Calculate the weighted super-matrix. Combine all the sorting total assets or some assets of the enterprise. The purpose of en-
vectors that interact with each other at the network layer to terprise management is to effectively use various assets to obtain
form a super-matrix W under the control element GS. The the maximum profit. Profits are mainly derived from operating
weighted super-matrix is obtained by normalizing each income, and firms must rely on assets and make use of it to obtain
column, and the sum of the elements of each column in the operating income. The higher the asset turnover rate, the higher the
weighted super-matrix is 1. operation efficiency and the greater the profit. By analyzing the
(4) Calculate the limit super-matrix. That is, the weighted super- efficiency of asset utilization, it is possible to evaluate whether the
matrix is solved by the of K power, and the absolute ordering business revenue of the firm maintains reasonable relationship
K with various operating assets, and to examine the assets utilization
vector is calculated lim W . If the limit is convergent and
k/∞ efficiency of the enterprise. Typical indicators include: Margin of
unique, then the convergence result can be used as the in- safety ratio, Merchandise turnover and Cross ratio.
tegrated weight value for each element.
0 1 4.3. Debt
ðj1Þ ðj2Þ
/ ð jnj Þ
B wi1 wi1 wi1 C
B C Debt management is an important feature of modern firms.
B
Wij ¼ B wi2
ðj1Þ
wi2 ðj2Þ / wi2 ðjnj Þ C
C (7) While firms use debt to obtain the effect of liabilities, they also
B « « « « C
@ A amplify business risks and become financial risks. At the same time,
wini ðj1Þ wini ðj2Þ / wini ð jnj Þ
the liability efficiency is also constrained by factors such as agency
costs and financial constraints. Reasonable determination of the
degree of corporation debt is a problem that must be faced by firms
in debt management. The main indicators are: asset-liability ratio,
4. Evaluation index system for financial effciency
debt to equity ratio, tangible net debt ratio, and interest coverage
optimization
ratio (Wang et al., 2010).

This study selects the indicators for the financial efficiency

4.4. Asset management
optimization from the four aspects of Earning ability, efficiency
Debt management and Asset management based on the actual
The asset management ratio is a measure of the firms' financial
situation of the financial efficiency.
efficiency in asset management, which is an indicator of the firms’
asset turnover. Operational capability management is a general
4.1. Earning ability
term of informationization solutions of manufacturing and enter-
prise informationization for asset-intensive firms. It aims to
Earning ability is the ability of a firm to make a profit. The
improve asset availability and reduce operating and maintenance
earning ability ratio is a comparison of certain resources and output
costs of firms and optimizes enterprise maintenance resources as
(net) profits. Typical indicators are earnings per share, dividends
the core. Through information technology, reasonable arrange-
per share, price-to-earnings ratio, and net assets per share. In
ments for maintenance plans and related resources and activities. It
general, a firms’ profitability involves only normal business con-
is possible to increase revenue, and by optimizing the arrangement
ditions. Although the abnormal business conditions bring benefits
of maintenance resources, the cost can be reduced, thereby
or losses to the firm, this is only an individual result under special
improving the economic efficiency of the enterprise and the market
circumstances and cannot explain the earning ability of the firm
competitiveness of the enterprise by increasing the availability of
(Ohlson and Juettner-Nauroth, 2005). This study excludes when
equipment (Malkiel, 2013). The main indicators included are: in-
analyzing corporation earning ability.
ventory turnover ratio, accounts receivable turnover ratio, oper-
ating cycle.
a. Non-normal items such as securities trading;
In summary, this study establishes a relatively complete finan-
b. Business projects that have been or will be stopped;
cial efficiency optimization index system, showed in Table 2. The
c. Special items such as major accidents or legal changes;
distinguishing feature of this indicator system is that the peer in-
d. Factors such as the cumulative impact of changes in accounting
dicators are not relatively independent, but rather present a logical
standards and financial systems.
relationship of mutual influence and interdependence.

Table 1 5. Application case of financial effciency evaluation model

Meaning of 1e9 scale.
5.1. Determine the relationship between the primary indicators
Importance Degree Scale
based on DEMATEL
i and j is of equal importance (or equivalent) 1
i is slightly more important than j 3
i is significantly more important than j 5 1 Calculate the direct relationship matrix
i is more important (or better) than j 7
i is extremely important than j 9 First, a group of four people will uses the Delphi method to
Median value of the above two adjacent judgments 2,4,6,8 judge the direct influence relationship between the primary in-
Median value of the above two adjacent judgments
dicators and construct a directed graph and transform the directed
6 S. Lan et al. / Journal of Cleaner Production 237 (2019) 117769

Table 2
Index evaluation system.

Secondary Indicator

U1 Earning Ability U11 Earnings per share

U12 Dividend per share
U13 Price earnings ratio
U14 Net assets per share
U2 Efficiency U21 Margin of safety ratio
U22 Merchandise turnover
U23 Cross ratio Fig. 3. Direct relation matrix X.
U3 Debt Management U31 Asset-Liability Ratio
U32 Debt to equity ratio
U33 Tangible net debt ratio
U34 Interest coverage ratio
U4 Asset management U41 Inventory turnover ratio
U42 Account receivable turnover
U43 Operating cycle

graph into the form of the matrix, which is the required direct
relation matrix A (as shown in Fig. 2). Fig. 4. Comprehensive influence matrix T.

2 Calculate the normalized direct relation matrix X by Equations

①,② as shown in Fig. 3. Table 3
3 On the basis of obtaining the standardized direct relation matrix Adjusted comprehensive impact matrix (retains 3 decimal points).
X, the comprehensive influence matrix T is obtained by using U1 U2 U3 U4 D DþR D-R
Equation (3) (Fig. 4).
U1 0.252 0.345 0.445 0.418 1.460 3.119 0.200
4 Constructing a causal map U2 0.300 0.217 0.393 0.300 1.210 2.888 0.468
U3 0.637 0.712 0.461 0.637 2.446 4.311 0.581
In order to further analyze the comprehensive influence rela- U4 0.471 0.404 0.567 0.305 1.747 3.406 0.087
tionship between the financial efficiency optimization factors, the R 1.659 1.678 1.865 1.659

influence degree D, the affected degree R and the centrality D þ R

and the cause degree D-R of each factor are calculated as shown in
Table 3. Where D represents the sum of the matrix row elements;
R represents the sum of the matrix column elements; D þ R
represents the position of an indicator in all indicators and the size
of its effect; if D-R is a positive number, it indicates that the in-
fluence of the indicator on other indicators is large, which is called
the cause factor. If the D-R is negative, it means that the element is
greatly affected by other factors, which is called the result factor.
A causal map can be constructed according to D þ R, D-R, where
the arrow points to indicate the extent to which other indicators
affect it.
Fig. 5 presented that U3 (Debt Management) and U4 (Asset
management) are the causal factors, which have a large impact on
the indicators U1 (Earning ability) and U2 (Efficiency), and are the
most fundamental driving factors driving the optimization of

U1 U2 U3 U4
Fig. 5. Causal map of primary indicators.

U1 0 1 2 2 financial efficiency. U1 and U2 are the result factors, which are the
most direct driving factors affecting the optimization of financial
U2 1 0 2 1 efficiency. The more obvious the effect on optimizing the financial

A efficiency from the perspective of the centrality, the greater the

centrality. The effect of each level of indicators on the optimization

U3 3 4 0 3 of financial efficiency is U3, U4, U1, U2 from large to small. More

importantly, through Fig. 5, this study can clearly identify the
impact relationship between each level of indicators, and each level
has internal dependencies, which lays the foundation for the
U4 2 1 3 0 application of ANP method.
Fig. 2. Direct relation matrix A.
 S. Lan et al. / Journal of Cleaner Production 237 (2019) 117769 7

5.2. Calculate the relevant weights of indicators at all levels based

on fuzzy ANP

The weight of the indicators will be calculated by applying the

ANP method after determining the relationship between indicators
through the DEMATEL method.

5.2.1. Building a network structure

The primary indicator causal map generated by the DEMATEL
method and combined with the actual situation in the financial
efficiency optimization process, this study draws the ANP network
structure with only the target layer in the control layer (Fig. 6) (see
Fig. 8) (see Fig. 7).

5.2.2. Establish a judgment matrix

The 20 experts were randomly selected. The degree of mutual
influence between the element sets and elements of the pairwise
comparison was judged to construct the judgment matrix accord-
ing to the 1e9 judgment scale., there are many judgment matrices
to be constructed, and the calculation amount is larger since this
study involves 4 element sets and 14 elements. This study uses the
super deicision software to input the judgment matrix of the two
comparisons into the software for calculation in the questionnaire
mode. And the consistency check of the judgment matrix. Table 4
Fig. 7. Unit root test chart.
shows the factor judgment matrix in the U3 element set (Debt
Management) under the element U11 (Earnings per share) criterion.
According to the judgment matrix, the unweighted weight value of

Fig. 6. ANP network structure.

8 S. Lan et al. / Journal of Cleaner Production 237 (2019) 117769

Fig. 8. Impulse response of Earning, Efficiency, Debt and Asset.

Table 4 (0.047)>U13 price-earnings ratio (0.042)>U23 cross ratio (0.042)>

Judgment matrix of U3 element set under the U11 criterion. U11 earnings per share (0.028)>U12 dividend per share (0.027)> U21
U11 U31 U32 U33 U34 Weight operating safety ratio (0.008).
U31 1 4 3 1 0.105
U32 1/4 1 1/3 1/4 0.528 5.3. Empirical analysis of firm operation data
U33 1/3 3 1 1/3 0.262
U34 1 4 3 1 0.105
C.R. ¼ 0.0312(Consistency is less than 0.1)
The 14 operating indicators data of 41 firms from 2008 to 2017
were selected in order to objectively measure the impact of these
indicators on optimizing the financial efficiency. The entropy
weight method is used to determine the weight of each operational
each element in the element set U3 can be obtained, and the con-
indicator in each year, so as to obtain the average entropy weight
sistency check result is 0.0312 < 0.1, which satisfies the require-
reflecting the importance of the secondary indicator and compare it
ment, and the weight is within an acceptable range.
with the normalized secondary indicator importance index value
and important order based on DEMATEL-ANP. In order to further
5.2.3. Constructing a limit super-matrix obtain the average entropy weight reflecting the importance of the
Select the elements to be judged in turn, repeat the construction primary indicator and compare and test with the primary indicator
of the judgment matrix and determine its weight. Through calcu- importance index value and important order based on DEMATEL-
lation, the unweighted matrix, the weighted super-matrix and the ANP, the 14 indicators per year are combined by weighted sum-
limit super-matrix can be obtained. mation. And again, use the entropy weight method to obtain the
entropy weight and average entropy weight of the four primary
(4) Determine the weight of the indicator indicators. In order to further test the influence relationship be-
tween the primary indicators, based on the synthesized four-level
According to the results of the limit super-matrix, the weights of index data of 2008e2017 obtained above, the PVAR model was
each evaluation index are determined (as shown in Table 4). established, and the mutual influence relationship between them
Table 5 presented the most important element set for optimizing was analyzed by the impulse response method.
the financial efficiency is U3 Debt management (0.366), U4 Asset
management (0.363), U1 Earning ability (0.144), and U2 Efficiency 5.3.1. The importance of assessing primary and secondary
(0.127), which is consistent with the order of the primary indicators indicators based on entropy weighting
obtained in 3.1 based on DEMATEL's scoring data for experts. The entropy method determines the weight of the indicator
The weight ordering of each element can also be obtained from based on the amount of information provided by the observations
Table 4 in addition to the weight ordering of the element sets. The of each indicator. The entropy method is a method for determining
elements that play an important role in optimizing the financial the weight of each index according to the amount of information
efficiency are: U42 Account Receivable Turnover (0.152)>U41 In- provided by the indicator data. The idea is that the larger the dif-
ventory turnover ratio (0.144)>U34 Interest coverage ratio (0.118) ference between the values of the evaluation objects on a certain
>U31 asset liability ratio (0.108))> U32 debt to equity ratio (0.078) indicator, the more important the weight is correspondingly, the
>U22 merchandise turnover ratio (0.087)>U43 operating cycle weight of the indicator is determined according to the influence of
(0.067)>U33 tangible net debt ratio (0.062)>U14 net assets per share the relative change degree of the index on the system as a whole,
 S. Lan et al. / Journal of Cleaner Production 237 (2019) 117769 9

Table 5
Indicator weights for financial efficiency optimization.

Primary indicator Secondary indicator A B C D E F Average

U1 Earning ability (0.144) U11 Earnings per share 0.029 0.027 0.028 0.029 0.027 0.029 0.028
U12 Dividend per share 0.027 0.027 0.027 0.028 0.027 0.027 0.027
U13 Price earnings ratio 0.044 0.042 0.042 0.041 0.041 0.042 0.042
U14 Net assets per share 0.047 0.045 0.049 0.047 0.047 0.046 0.047
U2 Efficiency (0.127) U21 Margin of safety ratio 0.008 0.007 0.010 0.008 0.007 0.007 0.008
U22 Merchandise turnover 0.077 0.075 0.075 0.075 0.078 0.083 0.077
U23 Cross ratio 0.042 0.043 0.041 0.042 0.043 0.042 0.042
U3 Debt management (0.366) U31 Asset-Liability Ratio 0.109 0.108 0.108 0.108 0.109 0.106 0.108
U32 Debt to equity ratio 0.079 0.077 0.078 0.079 0.078 0.077 0.078
U33 Tangible net debt ratio 0.063 0.062 0.063 0.062 0.063 0.060 0.062
U34 Interest coverage ratio 0.118 0.120 0.119 0.118 0.118 0.115 0.118
U4 Asset management (0.363) U41 Inventory turnover ratio 0.142 0.144 0.146 0.145 0.143 0.146 0.144
U42 Account receivable turnover 0.154 0.151 0.152 0.151 0.154 0.150 0.152
U43 Operating cycle 0.063 0.073 0.063 0.066 0.066 0.070 0.067

and the index with a relatively high degree of change has a large 5.3.1.4. (4)Calculating the Entropy Value ej under the jth Indicator.
weight. This method is widely used in various fields such as sta- The entropy value ej under the jth indicator can be worked out.
tistics and assessment (Reesor, 2001). According to the degree of mrepresents the number of evaluated objects. When pij ¼ 0 or
variation of each indicator, the weight of each indicator can be pij ¼ 1, pij  lnðpijÞ ¼ 0.
calculated objectively, which provides a basis for comprehensive
evaluation of multiple indicators. The entropy method has strong 1 X m
ej ¼  pij  lnðpijÞ (11)
objectivity and avoids the error caused by the subjective judgment lnðmÞ
i¼1
of the evaluator (Hendi and Sheykhi, 2011).
The analysis of firm operation data based on entropy weight
method is divided into the following steps:
5.3.1.5. (5)Calculating the difference coefficient dj under the jth in-
5.3.1.1. (1)Decision Matrix Setting up. There are m firms partici- dicator. Observe the formula of entropy value, for a certain indi-
pating in the evaluation, and the firms’ operating status has n cator dj, the more similar the variances of vij are, the higher the
evaluation indicators, so the evaluation system consists of m value of ej. Difference coefficient dj can be determined according to
samples and n evaluation indicators. Then form a decision matrix X: the definition of entropy. The high value of dj means the more in-
2 3 formation the corresponding indicator provides, as a result, more
6 x11 x12 ::: x1n 7 weighting should be given to the indicator.
6 x12 x22 ::: x2n 7
X¼6
6 «
7 (8) dj ¼ 1  ej
4 « « « 75
(12)
xm1 xm2 ::: xmn

5.3.1.6. (6)Determining the weightings for each indicator. Then

calculate the entropy weight of each indicator. n represents the
5.3.1.2. (2)Standardized Decision Matrix。. Because the objectivity
number of indicators. Results are shown as follows in Table 6:
and negativity, measurement units of each indicator are different, it
is unrealistic to use the data directly for calculation. Through dj
normalization of the decision matrix X to form a standardized wj ¼ j ¼ 1; 2; :::; n
P
n (13)
matrix V ¼ ðvijÞm  n. Among them, vij is the normalized value of dk
k¼1
Xij. Max (xj)and minðxjÞ are the maximum and minimum values of
the jth indicator. For each year's operational data above (cross-sectional data), the
entropy weight of 14 indicators can be obtained by using the above
xij  minðxjÞ entropy weight method. Finally, the entropy weight of each year is
vij ¼ (9)
maxðxjÞ  minðxjÞ shown in the following table, and the average entropy weight of the
indicator can be seen. The weights and the order of the normali-
zation of the indicators obtained by the ANP analysis are basically
5.3.1.3. (3)Weighting of the ith Evaluated Object under the jth Indi- the same. Through the actual firm operation data, it is feasible to
cator. For a certain indicator j, the bigger the variance of vij means the explain the importance of the DEMATEL-ANP method to determine
more influence the indicator has on the evaluated object, namely the importance of the secondary indicators.
more useful information are provided to measure evaluated object by In order to further test the weight of the primary indicator
the indicator. Entropy can be used to measure how much information determined by the DEMATEL-ANP method, the second-level indi-
are provided according to the definition of entropy. Under the jth cator included in the primary indicator is weighted for each year's
indicator, weighting of the ith evaluated object is pij 0  pij  1. data, and the weight is the above-mentioned entropy right, and the
firm can be obtained in the year. The scores of the primary in-
vij dicators are detailed in Table 7, and the entropy weight method is
pij ¼ (10)
P
m used again to obtain the annual entropy weight and average en-
vij tropy weight of the four primary indicators. The results are shown
i¼1
in the table below. The weights of the normalized primary in-
dicators obtained by the two methods are very close, and the order
10 S. Lan et al. / Journal of Cleaner Production 237 (2019) 117769

Table 6
Entropy weights of 14 secondary indicators for 2008e2017.

U11 U12 U13 U14 U21 U22 U23 U31 U32 U33 U34 U41 U42 U43

2017 0.009 0.012 0.017 0.062 0.021 0.141 0.101 0.016 0.081 0.050 0.085 0.140 0.192 0.074
2016 0.015 0.014 0.013 0.048 0.021 0.070 0.048 0.012 0.189 0.034 0.192 0.119 0.161 0.064
2015 0.009 0.009 0.014 0.027 0.006 0.061 0.038 0.015 0.400 0.035 0.111 0.096 0.132 0.047
2014 0.019 0.018 0.005 0.036 0.005 0.088 0.045 0.029 0.016 0.056 0.333 0.128 0.167 0.054
2013 0.023 0.023 0.030 0.071 0.022 0.105 0.046 0.029 0.078 0.061 0.073 0.178 0.201 0.061
2012 0.042 0.040 0.223 0.050 0.005 0.079 0.027 0.027 0.090 0.052 0.005 0.150 0.162 0.049
2011 0.055 0.055 0.017 0.057 0.008 0.101 0.053 0.050 0.013 0.079 0.006 0.207 0.221 0.079
2010 0.026 0.026 0.043 0.045 0.005 0.085 0.047 0.045 0.006 0.077 0.173 0.186 0.166 0.069
2009 0.029 0.011 0.011 0.038 0.005 0.093 0.036 0.043 0.007 0.075 0.095 0.168 0.159 0.230
2008 0.017 0.017 0.014 0.025 0.004 0.073 0.033 0.104 0.018 0.086 0.225 0.134 0.181 0.070
Weight by entropy 0.024 0.023 0.039 0.046 0.010 0.090 0.047 0.037 0.090 0.060 0.130 0.151 0.174 0.080
Weight by ANP 0.028 0.027 0.042 0.047 0.008 0.077 0.042 0.108 0.078 0.062 0.118 0.144 0.152 0.067

is consistent. The order of importance from high to low is U3, U4, U1, Table 8
U2 (see Table 8). Result of Hadri Lagrange multiplier stationarity test.

Variable Difference order Z stationarity

5.3.2. Investigating the influence relation between the primary Earning ability 0 0.7130 6.9603*** ✓
indicators based on the PVAR model (0.2379) (0.0000)
In order to further test the influence relation between the pri- 1 3.0480 1.9939 ✓
(0.9988) (0.9769)
mary indicators, based on the panel data of the four primary in-
Efficiency 0 5.1670*** 5.4724*** 
dicators of 41 firms from 2008 to 2017 obtained above, the data (0.0000) (0.0000)
stability test was first carried out to identify the four variables. The 1 0.0907 2.5355 ✓
stationary order is then studied using PVAR and investigated by the (0.5361) (0.9944)
Debt management 0 5.8899*** 3.0231 
impulse response function and variance decomposition.
(0.0000) (0.0013)
1 2.4352 1.3622 ✓
5.3.2.1. Stationarity test. (0.9926) (0.0866)
Asset management 0 6.3652*** 2.6756 
yit ¼ ri yit1 þ εit (14) (0.0000) (0.0037)
1 3.2922 1.3457 ✓
If jri j < 1, the sequenceyt is said to be wide (trend) stationary, (0.9995) (0.9108)
and if jri j ¼ 1, the sequenceyt includes the unit root, which is an notes: P values in parentheses, *, ** and *** indicate significant significance at 1%,
unstable sequence. There are many methods for stationarity test. 5%e10%,.
There are two common classification methods. One is based on
whether the individual auto-regressive coefficients of the cross-
section are the same in the test. It can be divided into two cate- efficiency debt and asset) is used as Y, that is,Yt ¼
gories: one type assumes that all individuals have the same auto- 0 1
regressive coefficient, that is, all the cross sections, that is, i B earningt C
haveri ¼ r. In this study, the Hadri Lagrange multiplier stationarity B Ddebtt C
Yt ¼ B
B
C
C (15)
test is performed on four variables (Hadri, 2000). The original hy- @ Defficiencyt A
pothesis is that there is no unit root, that is, the variable is stable. Dassett
The test results are shown in the table below. Earning ability is
stable on the horizontal value, while the first-order difference of Then, according to the information criterion, the lag order is
variables Efficiency, Debt management and Asset management are selected to construct the PVAR model.The results of the informa-
stable on the horizontal value. tion guidelines are shown in the table, whether it is MBIC, MAIC or
MQIC indicators, Y is a first-order lag, and the PVAR(1) model is:
5.3.2.2. PVAR model. Based on the above-mentioned panel data
stability analysis, the combination of the four variables of (earning, Yt ¼ rYt1 þ εt (16)

Table 7
Entropy weights for the four primary indicators in 2008e2017. 5.3.2.3. Unit root test of PVAR model. In order to further test
whether the PVAR model is stable (for the stationary process), the
U1 U2 U3 U4
unit root test is used, and the results are shown in the following
2017 0.088 0.656 0.085 0.170 table and the following figure. The four unit roots of the model fall
2016 0.183 0.226 0.197 0.394
2015 0.165 0.205 0.349 0.281
within the unit circle, so the model is stable.
2014 0.136 0.152 0.433 0.279
2013 0.217 0.130 0.346 0.307 5.3.2.4. Impulse response analysis. To describe the effect of unit
2012 0.214 0.138 0.386 0.261
variations of one variable in the model on other variables in the
2011 0.190 0.206 0.400 0.204
2010 0.296 0.147 0.360 0.197 system, based on the existing PVAR(1) model, an impulse response
2009 0.341 0.169 0.323 0.166 model (IRF) is used to establish an impulse response analysis model
2008 0.400 0.150 0.268 0.183 in order to study the response of other variables in the later stage
Weight by Entropy 0.223 0.218 0.315 0.244 after four variables are affected by one unit standard deviation.
Weight by ANP 0.144 0.127 0.366 0.363
Each row in the figure below shows the effect of the unit change of
 S. Lan et al. / Journal of Cleaner Production 237 (2019) 117769 11

one variable on other variables. Each column indicates that a var- and optimize the firms’ asset structure.
iable is affected by the unit change of other variables.
After the earning index was affected by a standard deviation of 5.4. Discussions and findings
efficiency at the initial stage, the index gradually increased from the
zero, and the positive reaction effect was strengthened from the Based on DEMATEL, four primary indicators have been estab-
beginning to the second period, rising to 0.0089 at the end of the lished, and the order of influence of these four indicators on the
first period when the instantaneous effect of the second period quality of financial efficiency is: U3, U2, U1, U4. Based on the ANP
reached its maximum. This positive effect will show a downward network structure, combined with the expert scoring data, the
trend after the second phase. When affected byDasset, the index order of weights of the primary indicators is consistent with the
earning gradually rises from the value of 0 and reaches the order of influence of DEMATEL's causal analysis on the quality of
maximum value of 0.011 in the second phase, and then gradually the financial efficiency and financial sustainability.
shows a downward trend, and the positive effect gradually ap- Based on the actual operational data of the 14 indicators gath-
proaches zero. ered from 41 firms in 2008e2017, the entropy weight method is
The earning impulse response by the Ddebt reaches the used to objectively test the relationship among these indicators and
maximum negative effect of 0.012 in the first phase, after which the impact of these indicators on the quality of financial operation
the negative effect will show a downward trend but still with a efficiency. The average weight of each of the 14 indicators. The
negative effect until the value reaches zero after the forth phase. order of weights of the indicators obtained from the ANP analysis is
The positive effect after the fourth period is not obvious, close to consistent. The weighted scores of the four primary indicators of 41
zero value. In general, the rising indicators of efficiency and assets firms in 2008e2017 are consistent with the order of the primary
will lead to a short-term increase in the firms’ earning indicator. But indicators based on DEMATEL's scores for expert scoring data and
there can be a negative effect on the earning indicators when it is the order of weights of the primary indicators obtained by ANP
affected by the debt in short-term, and the positive effect will not analysis. The above points verify the effectiveness of the DEMATEL-
be obvious in the long run. ANP method. The model analyzes the relationship between the
After the initial positive impact on earning and Dasset on the influencing factors in the firms' financial management, extracts the
standard deviation, the indexDefficiency shows a mild upward cause indicators from it, and provides theoretical support and
trend with the positive effect on the indicator at the early stage, and starting point for controlling financial risks.
both of them have the largest positive effect in the first period, The PVAR model is constructed, and the causal relationship
which are 0.012 and 0.013. After that, this positive effect shows a between the primary indicators is tested by impulse response
trend of decreasing gradually. In the long run, the increase in the analysis to reflect the dynamic impact between the indicators
firms' asset management and profitability indicators can decrease based on the score data of the primary indicators. The results show
operational efficiency. Similarly, when the efficiency has been that:
affected by the indicatorDdebt, the trend of maintaining the initial
period is to maximize the positive effect in the first period, which is 1) The asset management capability (Dasset) of the first phase of
0.015. After that, the positive effect is gradually decreased, and the lag has a positive and significant impact on the firms' profit-
second stage is the fastest. It shows that the increase of the firms’ ability and efficiency in short term. At the same time, from a
debt index will also enhance operational efficiency to a certain long-term perspective, the improvement of asset management
extent and cannot increase firm efficiency simply by relying on the will lead to higher liability of business operations.
reduce of the corporation debt in the short term. 2) Following the initial impact of the standard deviation, the in-
At the beginning, Ddebt showed a negative effect after it was dicator Ddebt has a negative effect on the earning indicator in
positively impacted by earning, and Dasset. Among them, when it the short term. But in the long run, the high debt ratio is
was affected by earning, Ddebt showed a negative effect in the negatively related to the firms' asset management and opera-
initial stage and reached a maximum of 0.039 in the first period, tional efficiency.
after which the negative effect showed a trend of increasing 3) The indicatorDefficiency exert a positive influence on earning in
gradually, and gradually came close to zero in the long run. The the short term and the earning indicator shows a trend of rising
indicator Ddebt began to show a trend of increasing from the first and then decrease gradually, but the value is still above 0. In
beginning figure 0.061 value but still has a positive correlation the long run, the improvement of the firms' operational effi-
after it was influenced by the Defficiency in the long-term. When ciency has reduced firms' asset management and profitability.
affected by the Dasset; the indicator Ddebt continues to decrease At the same time, the improvement of operational efficiency and
from the 0 value, and the maximum instantaneous negative effect the firms' debt level have a positive correlation.
in the first period is 0.062, after which the negative effect de- 4) Following the initial impact of the standard deviation of indi-
creases gradually and tends to zero. This shows that the asset cator earning, the indicator Dasset has a significant upward
management and the ability to earn profits of a company are trend in the early stage, and the indicator Defficiency also shows
negatively correlated with high debt rate in the short run. But in the a significant positive effect in the short run. It shows that there is
long term, they have positive correlation. a positive correlation between profitability indicators and asset
After the index Dasset was positively impacted by a standard indicator in the short term. The rise of profit indicators lead to
deviation of Defficiency at the beginning, it began to show a trend the decrease of efficiency and asset management indicators in
of decreasing from the beginning at the highest value of 0.049, and the long run. However, for the debt indicator, there is a negative
gradually approached the value of 0 in the sixth period. The indi- effect at the beginning, and in the long run, the two are posi-
cator Dasset showed a similar trend as the impulse response of the tively correlated.
indicatorDdebt and earning: The Dasset indicator reaches the
positive instantaneous effect at the first stage, the maximum of
0.007 and 0.004. The downward trend will gradually approach zero References
in the long run. It shows that the increase of the liabilities and firms'
Martínez-Ferrero, J., Frías-Aceituno, J.V., 2015. Relationship between sustainable
profitability will have a negative impact on firms' asset in the long development and financial performance: international empirical research. Bus.
term, but in the short run it will help improve the firms' asset level Strateg. Environ. 24 (1), 20e39.
12 S. Lan et al. / Journal of Cleaner Production 237 (2019) 117769

Afonso, A., Jalles, J.T., 2016. The elusive character of fiscal sustainability. Appl. Econ. approach: the case of shipping operations. J. Clean. Prod. 105, 275e284.
48 (28), 2651e2664. Lam, J.S.L., Lai, K., 2015b. Developing environmental sustainability by ANP-QFD
Benitez, J., Chen, Y., Teo, T.S.H., et al., 2018. Evolution of the impact of e-business approach: the case of shipping operations. J. Clean. Prod. 105, 275e284.
technology on operational competence and firm profitability: a panel data Malkiel, B.G., 2013. Asset management fees and the growth of finance. J. Econ.
investigation. Inf. Manag. 55 (1), 120e130. Perspect. 27 (2), 97e108.
Bratten, B., Causholli, M., Omer, T.C., 2019. Audit firm tenure, bank complexity, and Martellini, L., Milhau, V., Tarelli, A., 2017. Capital structure decisions and the optimal
financial reporting quality. Contemp. Account. Res. 36 (1), 295e325. design of corporate market debt programs. J. Corp. Financ. 49, 141e167.
Büyüko €zkan, G., Güleryüz, S., Karpak, B., 2017. A new combined IF-DEMATEL and IF- S0929119917307198.
ANP approach for CRM partner evaluation. Int. J. Prod. Econ. 191, 194e206. Martínez-Campillo, A., Fern andez-Santos, Y., del Pilar Sierra-Ferna ndez, M., 2018.
Cantele, S., Zardini, A., 2018. Is sustainability a competitive advantage for small How well have social economy financial institutions performed during the
businesses? An empirical analysis of possible mediators in the crisis period? Exploring financial and social efficiency in Spanish credit unions.
sustainabilityefinancial performance relationship. J. Clean. Prod. 182. J. Bus. Ethics 151 (2), 319e336.
Chae, H.C., Koh, C.E., Park, K.O., 2018. Information technology capability and firm Mavi, R.K., Standing, C., 2018. Critical success factors of sustainable project man-
performance: role of industry. Inf. Manag. 55 (5), 525e546. agement in construction: a fuzzy DEMATEL-ANP approach. J. Clean. Prod. 194,
Checherita-Westphal, C., Hughes Hallett, A., Rother, P., 2014. Fiscal sustainability 751e765.
using growth-maximizing debt targets. Appl. Econ. 46 (6), 638e647. Navarro-Galera, A., Rodríguez-Bolívar, M.P., Alcaide-Mun ~ oz, L., et al., 2016.
Chen, R.H., Lin, Y.H., Tseng, M.L., 2014. Multi-criteria analysis of sustainable devel- Measuring the financial sustainability and its influential factors in local gov-
opment indicators in the construction minerals industry in China. Resour. Policy ernments. Appl. Econ. 48 (41), 3961e3975.
46, 123e133. Nguyen, H.T., Dawal, S.Z.M., Nukman, Y., Rifai, A.P., Aoyama, H., 2016. An integrated
Cheng, M., Dhaliwal, D., Zhang, Y., 2013. Does investment efficiency improve after MCDM model for conveyor equipment evaluation and selection in an FMC
the disclosure of material weaknesses in internal control over financial based on a fuzzy AHP and fuzzy ARAS in the presence of vagueness. PLoS One 11
reporting? J. Account. Econ. 56 (1), 1e18. (4), e0153222.
Coad, A., Rao, R., 2008. Innovation and firm growth in high-tech sectors: a quantile Ohlson, J.A., Juettner-Nauroth, B.E., 2005. Expected EPS and EPS growth as de-
regression approach. Res. Policy 37 (4), 633e648. terminants of value. Rev. Account. Stud. 10 (2e3), 349e365.
Curtis, A., Lewis-Western, M.F., Toynbee, S., 2015. Historical cost measurement and Reesor, R.M., 2001. Relative Entropy, Distortion, the Bootstrap and Risk.
the use of DuPont analysis by market participants. Rev. Account. Stud. 20 (3), Taqi, M., Ajmal, M., Ansari, M.S., 2018. Financial efficiency of India tourism devel-
1210e1245. opment corporation (ITDC) limited: an empirical study. J. Tour. Manag. Res. 5
Edmans, A., Heinle, M.S., Huang, C., 2016. The real costs of financial efficiency when (1), 14e22.
some information is soft. Rev. Financ. 20 (6), 2151e2182. Tseng, M.L., Tan, K., Chiu, A.S.F., 2016. Identifying the competitive determinants of
Fouejieu, A., Popescu, A., Villieu, P., 2019. Trade-offs between macroeconomic and firms' green supply chain capabilities under uncertainty. Clean Technol. Envi-
financial stability objectives. Econ. Modell. (Article in Press). ron. Policy 18 (5), 1247e1262.
Gabus, A., Fontela, E., 1972. World Problems, an Invitation to Further Thought Tseng, M.L., Wong, W.P., Soh, K.L., 2018b. An overview of the substance of resource,
within the Framework of DEMATEL. Battelle Geneva Research Center, Geneva, conservation and recycling. Resour. Conserv. Recycl. 136, 367e375.
Switzerland, pp. 1e8. Tseng, M.L., Phan, A.T., Lin, C.W., Jeng, S.Y., Negash, Y., Darsono, S.N.A., 2019a. Sus-
Ga^rleanu, N., Pedersen, L.H., 2018. Efficiently inefficient markets for assets and asset tainable Investment: interrelated among corporate governance, economic
management. J. Financ. 73 (4), 1663e1712. performance and market risks using investor preference approach. Sustain. J. 11,
Giordani, P., Jacobson, T., Schedvin, E.L.V., et al., 2011. Taking the twists into account: 2108.
predicting firm bankruptcy risk with splines of financial ratios. SSRN Electron. J. Tseng, M.L., Wu, K.J., Ma, L., Kuo, C.K., Sai, F., 2019b. A hierarchical framework for
49 (4), 1071e1099. assessing corporate sustainability performance using a hybrid fuzzy synthetic
Govindan, K., Shankar, K.M., Kannan, D., 2016. Application of fuzzy analytic network method-DEMATEL. Technol. Forecast. Soc. Chang. 144, 524e533.
process for barrier evaluation in automotive parts remanufacturing towards Wang, X., Dennis, L., Tu, Y.S., 2010. Measuring financial condition: a study of U.S.
cleaner productionea study in an Indian scenario. J. Clean. Prod. 114, 199e213. states. Public Budg. Finance 27 (2), 1e21.
Grewatsch, S., Kleindienst, I., 2017. When does it pay to be good? Moderators and Wang, H., Huang, J., Yang, Q., 2019a. Assessing the financial sustainability of the
mediators in the corporate sustainabilityecorporate financial performance pension plan in China: the role of fertility policy adjustment and retirement
relationship: a critical review. J. Bus. Ethics 145 (2), 383e416. delay. Sustainability 11 (3), 883.
Hadri, K., 2000. Testing for stationarity in heterogeneous panel data. Econom. J. 3 Wang, C., Ghadimi, P., Lim, M.K., Tseng, M.L., 2019b. A literature review of sus-
(2), 148e161. tainable consumption and production: a comparative analysis in developed and
Hendi, S.H., Sheykhi, A., 2011. Entropic corrections to Einstein equations. Phys. Rev. developing economies. J. Clean. Prod. 206, 741e754.
D 83 (8), 084012. Wu, K.J., Tseng, M.L., Chiu, A.S.F., Lim, M.K., 2017. Achieving competitive advantage
Hirunyawipada, T., Xiong, G., 2018. Corporate environmental commitment and through supply chain agility: a closed-loop hierarchical decision-making
financial performance: moderating effects of marketing and operations capa- structure under uncertainty. Int. J. Prod. Econ. 190, 96e107.
bilities. J. Bus. Res. 86, 22e31. Yang, C., Lan, S.L., Tseng, M.L., 2018. Coordinated development path of metropolitan
Jha, A., 2011. Financial reports and social capital. J. Bus. Ethics (1), 1e30. logistics and economy in Belt and Road using DEMATEL-Bayesian analysis. Int. J.
Kirschenmann, K., Malinen, T., Nyberg, H., 2016. The risk of financial crises: is there Logist. Res. Appl. 22 (1), 1e24.
a role for income inequality? J. Int. Money Financ. 68, 161e180. Zhang, Q., Xu, Z., Feng, T., et al., 2015. A dynamic stochastic frontier model to
Ko€lbel, J.F., Busch, T., Jancso, L.M., 2017. How media coverage of corporate social evaluate regional financial efficiency: evidence from Chinese county-level panel
irresponsibility increases financial risk. Strateg. Manag. J. 38 (11), 2266e2284. data. Eur. J. Oper. Res. 241 (3), 907e916.
Lam, J.S.L., Lai, K., 2015a. Developing environmental sustainability by ANP-QFD