Construction of Financial Performance Evaluation System based on
Principal Component Analysis Algorithm and Its Application in Digital
Transformation Enterprises
SUJING GUO
Lyceum of the Philippines University Manila Campus,
Manila 1002,
PHILIPPINES
Abstract: - In the context of a strong national push toward the growth of the "digital economy", traditional
manufacturing companies are increasingly turning to new digital technologies for their digital transformation.
This paper aims to investigate the suitability of using a financial performance evaluation system for assessing
the success of digital transformation strategies employed by these enterprises. The application of principal
component analysis in digital transformation enterprises involves repeatedly selecting the main indicators in the
financial performance evaluation index system of manufacturing enterprises. Finally, a financial performance
evaluation index system suitable for analyzing digital transformation enterprises is constructed. The differences
in financial performance before and after transformation are analyzed, and a comprehensive evaluation and
comparative analysis are conducted on the financial performance of digital transformation enterprises and non-
digital transformation enterprises. The experimental results show that the average growth rate of total assets of
enterprises is 7.07%. The average growth rate of operating revenue is 20.99%. The standard deviations are
17.42% and 235.9%. There is a significant difference between the maximum and minimum values of these two
indicators, indicating that the average dispersion of these two indicators is relatively high. In the initial phases
of digital transformation implementation, enterprises that adopt digital technology experience a certain level of
profitability improvement, as shown by the results. Compared to businesses that have not undergone a digital
transformation, digitally transformed enterprises possess greater advantages and flexibility in digital operations.
Digital transformation has important theoretical and practical value in improving the financial management
level of digital transformation enterprises.
Key-Words: - Principal Component Analysis; Digital Economy; Enterprise Transformation; Enterprise
Competitiveness; Evaluation System.
Received: March 15, 2023. Revised: August 29, 2023. Accepted: September 28, 2023. Available online: November 11, 2023.
1 Introduction
Currently, the Chinese economy has shifted from
rapid growth to high-quality development, in which
the manufacturing industry, serving as the main
body, has become the primary propeller of long-
term stable growth and sustained, rapid economic
progress, [1]. Traditional manufacturing enterprises
are proactively responding to the digital strategy by
utilizing new digital technologies to achieve digital
transformation (DT). Manufacturing is crucial for
China's economic development, and a new
generation of digital technology serves as the core
driving force for its transformation and upgrading
efforts, [2], [3]. In the digital economy era,
achieving successful transformation and
development of manufacturing enterprises greatly
relies on the effective integration of emerging
digital technologies and traditional manufacturing
practices, [4]. Promoting the integration of "digital +
intelligent manufacturing" is of great significance to
the country's high-quality development. At present,
countries around the world have realized that the DT
of enterprises is an inevitable choice of the times.
The integration of cutting-edge digital technology
into traditional enterprises is essential for their
future growth. To achieve this objective, a set of
national strategies has been developed under the
umbrella of "moving from manufacturing to smart
manufacturing", [5]. This shows that digitalization is
also a national strategy, and the DT of the
manufacturing industry through the use of new-
generation digital technologies is also the focus of
the state and society, [6]. The research objective is
to investigate whether "Internet plus manufacturing"
has a favorable impact on the high-quality growth of
manufacturing firms, the operational performance of
firms pre- and post-digital transformation, and the
WSEAS TRANSACTIONS on BUSINESS and ECONOMICS
DOI: 10.37394/23207.2024.21.14
Sujing Guo
E-ISSN: 2224-2899
153
Volume 21, 2024
financial contrasts between digital transformation
and non-digital transformation firms. After
conducting a quantitative analysis of the text, the
vocabulary of enterprise digital transformation
obtained can be used to describe the degree of
manufacturing enterprise digital transformation.
Using factor analysis and entropy weight methods,
this study investigates financial performance to
analyze the scientific nature of evaluation methods
for digital transformation. The aim is to provide
more scientific and objective reference opinions for
each evaluation subject.
2 Literature Survey
In the field of digitally transformed businesses,
many scholars have studied various aspects of it
over the years. [7], utilized the socio-technical
perspective of ETICS theory to construct and define
the proactive capacities of information technology
and socially encoded knowledge processes. These
capabilities yield business transformation processes
in the digital age. The authors discovered varying
results regarding the effects of mediation and direct
relationships. Socialized knowledge processes
directly influence the proactive capabilities of
information technology and the digital business
transformation of the company. Coded knowledge
processes successfully support the proactive
capabilities of information technology and provide
stronger support for the company's corporate IT/IS
strategy, providing excellent opportunities for coded
knowledge practices to improve the digital approach
to corporate business process transformation, [7].
[8], examined the impact of corporate DT on the
information environment. According to the findings,
implementing DT in enterprises led to a noteworthy
rise in analyst coverage and improved accuracy of
public information. However, there has been no
significant change in accuracy concerning private
information. The quality of information disclosure
and the content of stock price information are the
primary factors influencing the relationship between
them. Cyber-attacks, market competition, and social
media all impact this relationship, making DT a
promising avenue for investigation in emerging
capital markets.
[9], applied enterprise architecture to urban
digital transformation by developing an architecture
that addresses system alignment and data integration
issues in urban digital transformation. A qualitative
method was employed to assess the suggested
architecture. Data from a Norwegian municipality
served as a case study and was gathered through
interviews to authenticate the application of
Enterprise Architecture to urban service digitization
to emulate the digitization of e-mobility in a smart
city.
Many organizations are embarking on digital
transformation to prepare for the future. Digitally
transformed organizations must be prepared to deal
with unpredictable dynamics and ubiquitous
digitization. Such an organization must incorporate
the duality of exploitation and exploration and the
convergence of business and technology into its
organizational design. [10], presented a framework
based on DBS Bank's digital transformation journey
and provided new managerial insights for
strategically driving digital transformation.
Much of the academic and professional interest
in exploring DT and enterprise systems has focused
on the external forces of technologies or
organizations at the expense of internal factors.
Dilek and Babak explored employee digital literacy
as an organizational availability to capture the
contextual factors that underlie the location and use
of digital technologies. An evidence-based approach
to information systems practice was used by
examining the interaction between employee digital
literacy and employee technology in the use of
digital technologies. The interactive effect between
literacy and employee skills contributes to the new
concept of digital literacy available to organizations,
[11].
The Financial Performance Evaluation (FPE)
System aims to validate a direct approach to
measuring relational capital via corporate brand
estimation. Relationship capital management
impacts both financial performance and brand
development. Brand value serves as a reflection of
relationship capital. Based on empirical data, a
specific group of market and accounting metrics in
the IFRS framework presents crucial information for
assessing brand value. Altering the reference dataset
and model assumptions does not yield significant
alterations in the research findings, [12].
[13], introduced a new holdings-based procedure
to assess fund performance. Determining whether a
mutual fund's benchmark variance aligns with its
investment strategy is crucial. Funds that exhibit
benchmark discrepancies entail greater risks than
what is disclosed in their prospectus. Before further
risk adjustment, the funds on average outperformed
the prospectus benchmarks.
To assess the use of hybrid renewable energy
configurations in data center cooling units, [14],
examined the importance of free cooling technology
and compared it to the potential of renewable energy
systems. The data center consumed a large amount
of energy and the effectiveness of both methods in
WSEAS TRANSACTIONS on BUSINESS and ECONOMICS
DOI: 10.37394/23207.2024.21.14
Sujing Guo
E-ISSN: 2224-2899
154
Volume 21, 2024
various configurations was evaluated through
comparative analysis in terms of energy and water
savings, net present value, and emission rates. To
discover the maximum cooling energy-saving
potential of this data center, the combination of
these two methods was studied in the case of
Tehran.
[15], examined an organizational theory
approach to strategy, implementing performance
indicators to assess economic relations. They
developed a technique for comparative performance
evaluation that considered cooperation strategy and
identified economic performance as a crucial input-
output indicator for firms. Seven financial indicators
were selected, and timeliness was deemed essential
for a thorough economic assessment of a company.
After assessing the economic impact of 5G industry
growth, they concluded that the information and
telecommunication technologies were increasingly
becoming a new driver of economic growth, [15].
[16], analyzed the prediction model for competitive
dilemmas and discovered that there are inconsistent
rankings that correlate with different standards and
metrics of performance. To overcome this problem,
a multi-criteria decision support tool for predicting
corporate credit risk and distress has been proposed.
It provided multi-criteria evaluation for competitive
distress prediction models.
In summary, many scholars have launched
research on enterprise DT, and the application of the
FPE system in enterprise development is also very
extensive. However, most studies concentrate on
risk prediction models, and there is relatively little
research regarding the evaluation of the financial
and operational performance of digital enterprises
using principal component analysis. The study
chooses text mining and principal component
analysis (PCA) to construct an FPE system and
explore its application in DT enterprises.
3 Construction of FPE Index System
for Digital Transformation
Enterprises based on the PCA
Method
Various stakeholders in a company have their
concerns about the company's financial situation,
but a single financial indicator can only reflect one
aspect of the company. Thus, to ensure satisfactory
outcomes for various stakeholders within the
company, it is imperative to implement a
comprehensive and reasonable FPE index system
across different levels. This will enable a
comprehensive and integrated evaluation of the
organization's entire production and operational
processes.
3.1 Construction of Financial Performance
Evaluation Index System for DT
Enterprises
The inadequate comprehensiveness and timeliness
of the indicator system construction leads to
suboptimal performance evaluation, making it
essential to establish a scientifically and
comprehensively designed FPE indicator system for
digital transformation enterprises. This step is the
foundation and most crucial aspect of achieving
FPE within digital transformation enterprises, [17].
The method of combining qualitative analysis and
PCA is utilized to construct a preliminary indicator
system of 27 indicators for the FPE of
manufacturing enterprises. The system is based on
the dimensions of profitability, solvency,
development, and operational capacity, adhering to
the principles of relevance, systematicity,
importance, and feasibility. The FPE index system
of manufacturing enterprises is illustrated in Figure
1.
Dimension
Profitability
Solvency
Operating
capacity
Development
capability
●Return on assets
●Net profit margin of total assets
●Roe
●Return on invested capital
●Operating margin
●Ratio of Profits to Cost
●Earnings per share
●Current ratio
● Cash Current liability ratio
●Quick ratio
●Times interest earned
● Asset liability ratio
●Total asset turnover rate
●Current asset turnover rate
●Fixed asset turnover rate
●Inventory turnover
●Accounts receivable turnover rate
●Equity Turnover
●working capital turnover
●Total asset growth rate
●Sales expense growth rate
●Growth rate of management expenses
●Rate of capital accumulation
●Operating profit growth rate
●Net asset growth rate
●Operating revenue growth rate
●Net profit growth rate
●Capital accumulation rate
●Total operating costs
●Net assets per share
Fig. 1: FPE index system for manufacturing
enterprises
In the optimization of the indicator system, the
indicators are first screened, and the commonly used
methods are: PCA, conditional generalized variance
minimization method, great irrelevance method,
expert consultation method, [18], [19]. The study
carried out PCA on the initially selected financial
indicators according to different dimensions, to
establish a set of comprehensive and objective FPE
index systems. The financial indicators related to the
four dimensions of profitability, solvency,
development ability, and operating ability were
selected. Using PCA, the preliminary indicators
under each dimension are screened, to obtain the
representative indicators of each dimension. The
WSEAS TRANSACTIONS on BUSINESS and ECONOMICS
DOI: 10.37394/23207.2024.21.14
Sujing Guo
E-ISSN: 2224-2899
155
Volume 21, 2024
screening steps of the indicators are shown in Figure
2.
Determine the
next indicators
to be selected
Begin Standardization
of indicator data
Calculation of
correlation matrix
for standard data
Extracting and
Sorting Matrix
Eigenvalues
Extracting
eigenvalues and
eigenvectors
Calculation of
contribution rate
of eigenvalues
Determine the principal
component (contribution
value>1)
End
Fig. 2: Screening steps for indicators using principal
component analysis
The representative indicators in each of the four
selected dimensions were synthesized and subjected
to PCA. All the selected indicators underwent
multiple and repeated trial calculations and
screening, augmented by subjective judgment and
selection, to achieve the optimal FPE index system.
The representative indicator system selected in the
four dimensions is shown in Figure 3.
Profitability Solvency Operating
capacity
Development
capability
Return on assets
net profit rate
of total assets
Roe
Return on
invested capital
Ratio of Profits
to Cost
Quick ratio
Asset liability
ratio
Current asset
turnover rate
Total Asset
turnover
Current ratio
Operating
capacity
Fig. 3: A representative financial evaluation
indicator system in four dimensions
Figure 3 presents the results of the study's
analysis of seven profitability indicators to illustrate
PCA. The selected indicators accurately represent
the profitability dimension. Profitability is a
company's ability to generate earnings over a period
of time. Using PCA, after screening the initial
selection of indicators for the profitability
dimension, five representative indicators were
selected: return on assets, net profit margin on total
assets, return on net assets, return on invested
capital, and cost and expense margin. Asset return
rate refers to the ratio of the sum of net profit,
interest expense, and income tax of a company over
a certain period of time to the average total amount
of assets. The net profit margin of total assets refers
to the percentage of a company's net profit to the
average balance of total assets. Return on equity
(ROE) is a measure of a company's profitability
through investment over a certain operating period.
The return on investment capital is an indicator used
to evaluate the historical performance of a company,
mainly measuring the effectiveness of the invested
funds. Cost expense profit margin refers to the ratio
between a company's net profit and the total cost
expense. Solvency refers to the company's ability to
use its assets to repay short-term and long-term
debts. The quick ratio and gearing ratio were chosen
as the two indicators for the study. The quick ratio is
a representative indicator of a company's short-term
solvency, mainly representing the proportion of
quick assets in the company's current liabilities. The
debt-to-asset ratio to a certain extent represents the
size of a company's long-term debt repayment risk,
mainly reflecting the proportion of the company's
total liabilities to total assets. The company's
development ability pertains to its potential for
future expansion and the consequential changes in
business operations that will reflect the speed and
prospects for future growth. The study chooses two
representative indicators, namely the growth rate of
total assets and the operating income. The total asset
growth rate is a positive indicator that mainly
reflects the changes in the total assets of the
enterprise. It can provide more timely feedback on
changes in the enterprise's business strategy. The
growth rate of operating revenue is a direct
manifestation of a company's operating situation.
Compared to profits, operating revenue is less
affected by accounting and can reflect changes in
the company's operating situation more quickly. The
operational capability of an enterprise is the
operational management capability of an enterprise
in a particular operational cycle. On this basis, the
current asset turnover ratio and total asset turnover
ratio are proposed to represent the operational
capability of an enterprise in a particular period. The
turnover rate of current assets primarily indicates an
enterprise's capacity to utilize current assets for
generating operating income, thus serving as a
favorable indicator. The total asset turnover rate
refers to the ratio of a company's net operating
income to its total assets over a certain period of
time.
3.2 Comprehensive FPE of Digital
Transformation Enterprises based on
FA and Entropy Weight Method
When evaluating a company's financial performance
using the performance appraisal method, each
financial indicator's significance is determined. This
is achieved by assigning a power to each financial
indicator. The methods of assigning power to
WSEAS TRANSACTIONS on BUSINESS and ECONOMICS
DOI: 10.37394/23207.2024.21.14
Sujing Guo
E-ISSN: 2224-2899
156
Volume 21, 2024
financial indicators are categorized as shown in
Figure 4.
Fig. 4: Classification of methods for empowering
financial indicators
In Figure 4, it consists of subjective
empowerment and objective empowerment. Grey
correlation analysis, entropy weight method, PCA,
and FA are currently commonly used objective
empowerment methods. The study uses objective
empowerment methods such as PCA, FA, and
entropy weight methods to study the empowerment
of financial indicators and the evaluation method of
financial performance. The steps of the PCA method
are shown in Figure 5.
Matrix standardization
Correlation coefficient matrix
Eigenroots and
Eigenvector Matrices Calculation of principal
component contribution rate
Constructing a Comprehensive Scoring
Function for Enterprise Finance
Start
Calculate the principal
component load and the number
of principal components
End
Fig. 5: Steps of PCA
In the step of the PCA method in Figure 5, the
raw data matrix of the sample is shown in Equation
(1).
11 12 1
21 22 2
11
, ...
, ...
...
, ...
p
p
n n np
X X X
X X X
X
X X X







(1)
Equation (1),
Xij
is the
j
-th financial indicator
data of the
i
-th company to standardize the
indicator matrix as shown in Equation (2).
11 12 1
21 22 2
*
11
* , * ... *
* , * ... *
...
* , * ... *
p
p
n n np
X X X
X X X
X
X X X







(2)
The matrix of correlation coefficients is
calculated as shown in Equation (3).
11 12 1
21 22 2
11
, ...
, ...
...
, ...
p
p
n n np
r r r
r r r
R
r r r







(3)
The eigenroots and eigenvectors are calculated to
get the matrix as shown in Equation (4).
1
2
11
( )( )
( ) ( )
n
ki i kj j
k
ij nn
ki i kj j
kk
x x x x
r
x x x x



(4)
The contribution of principal components is
calculated to determine the principal components as
shown in Equation (5).
1
1
1
i
j
Zp
j
i
m
j
j
p
j
j
M
M











(5)
Equation (5),
M
represents the cumulative
contribution rate and
i
Z
M
stands for the
contribution rate of the principal component
i
Z
.
Principal component loadings can be calculated
based on the results of principal component analysis
and then combined with qualitative analysis to
determine their significance. Principal components
refer to a linear combination of the original financial
indicators. In this linear combination, the
coefficients of the individual variables are large or
small. They are both positive and negative. The
function for the composite score of a company's
financial performance is shown in Equation (6),
WSEAS TRANSACTIONS on BUSINESS and ECONOMICS
DOI: 10.37394/23207.2024.21.14
Sujing Guo
E-ISSN: 2224-2899
157
Volume 21, 2024
with the corresponding formula for FA also
presented in Equation (6).
1 1 2 2 ... ( 1,2,..., )
i i i im m i
X a F a F a F i p
(6)
In Equation (6),
12
, ,..., m
F F F
is the public factor.
i
is the special factor, and
im
a
is the factor loading
coefficient. The core of FA is to analyze the
correlation between numerous observed variables
through dimensionality reduction. The fundamental
structure of a significant volume of observational
data has been examined. Variables with a common
essence are grouped into a single factor to represent
the basic data structure with a limited number of
public factors. The method can extract the common
factors from the cluster of variables to explain the
structure among the variables at the cost of
minimum information loss. The flow of the FA
method used in the study is shown in Figure 6.
Applicability testing
Extracting
Common Factors
Varimax rotation
method
Calculation of various
factor variables
End
Start
Fig. 6: Factor analysis process
In Figure 6, the first step of FA is to evaluate its
applicability. If the factors are independent and the
correlation is not strong, common factors cannot be
obtained, rendering the FA unfeasible. So FA is
done to determine the correlation between the
original variables being analyzed and their
applicability. Generally, Bartlett's Spherical test and
KMO test are chosen for suitability analysis, [20].
The correlation matrix is tested by the Bartlett
Spherical test. If it is an identity matrix, the
observed data is not suitable for FA and vice versa.
Overall, at a significant level of < 0.05, it indicates
that there is a significant correlation in the original
variables and can be used for FA. KMO values
above 0.9 are most suitable for FA. KMO values
located in the middle of 0.7-0.8 are well-suited. 0.5-
0.7 are suitable. Values below 0.5 should be chosen
to be discarded. The study utilizes PCA for
extracting public factors. In determining the initial
factors, the eigenvalues, cumulative variance
contribution ratio of the factors, and fragmentation
diagram are examined. Furthermore, it is verified if
the selected principal components, which meet the
requirement of eigenvalues ≥ 1, encompass 85% of
the original data's information content. The formula
for the variance contribution ratio of the common
factors is shown in Equation (7).
1
( 1,2,..., )
i
i
Fp
k
k
M i p

(7)
Equation (7),
is the characteristic root of the
correlation coefficient matrix. The cumulative
contribution of public factors is calculated as in
Equation (8).
1
1
( 1,2,..., )
i
i
k
k
Fp
k
k
N i p

(8)
By rotating the factors, the properties of these
factors are reflected clearly. This helps to determine
the degree of influence each factor has on the others,
and subsequently identify which factors pertain to
each other. After the factor variables are
determined, the specific scores of each sample data
in each different factor need to be calculated. By
applying the FA method, the scores and rankings of
each public factor can be computed to determine the
factors that greatly influence the operation and
management of the company. This ability to
accurately pinpoint the entry point to enhance the
operation and management of the company is
extremely advantageous. The entropy weight
method is one of the objective assignment methods,
which can avoid the arbitrariness of manual
subjective judgment. Firstly, the indicators are pre-
processed to eliminate the gap between the
indicators, and the standardization formula for the
positive indicators is illustrated in Equation (9).
min( )
max( ) min( )
j i j
ij
i j ij
Xi X
YXX
(9)
The normalization formula for negative
indicators is illustrated in Equation (10)
max( )
max( ) min( )
j i j
ij
i j ij
Xi X
YXX
(10)
The fitness indicator's standardization is
illustrated in Equation (11).
1max( )
i j j
ij
i j j
XX
YXX

(11)
WSEAS TRANSACTIONS on BUSINESS and ECONOMICS
DOI: 10.37394/23207.2024.21.14
Sujing Guo
E-ISSN: 2224-2899
158
Volume 21, 2024
In Equation (11),
j
X
there are fixed values of
the moderation criteria. The characteristic weight of
the
j
-th indicator is calculated, i.e. contribution.
The characteristic ratio of the
i
-th enterprise is
calculated as shown in Equation (12).
1
ij
ij n
ij
i
Y
PY
(12)
For the
j
-th indicator, the entropy value is
calculated as shown in Equation (13).
1
1ln( ),0 1
ln
n
j ij j
i
e P e
n
(13)
The coefficient of variation is calculated by
Equation (14).
1
jj
ge
(14)
The weights of the evaluation indicators are
displayed in Equation (15).
1
, 1,2,...,
j
jm
j
i
g
W j m
g

(15)
4 Analysis of Evaluation Results and
Comparison of Digital and Non-
Digital Enterprises
For the 11 financial performance evaluation
standards selected by the paper, the PCA method is
applied. Seven indicators of profitability dimension
are used as examples to start the analysis.
Representative components are screened out. The
selected indicators undergo multiple rounds of trial
and error screening, supplemented by subjective
judgment and selection, ultimately obtaining an
ideal financial performance evaluation indicator
system.
4.1 Analysis of the Screening Results of FPE
Indicators based on PCA
On this basis, seven corporate profitability
indicators are selected as examples and subjected to
master meta-analysis. The 902 manufacturing
companies on the main board of A-shares in
Shanghai and Shenzhen are taken as the objects of
this study. The indicators of return on assets, net
profit margin on total assets, return on net assets,
return on invested capital, operating profit margin,
cost and expense margin, and earnings per share are
expressed as X1, X2, X3, X4, X5, X6, X7. Using
SPSS26.0 software, the correlation coefficient
matrix R and its eigenvalues between the indicators
are calculated. Table 1 displays the results.
Table 1. Matrix of correlation coefficients
X1
X2
X3
X4
X5
X6
X7
X
1
1
0.96
7
0.84
6
0.86
9
0.70
5
0.76
9
0.45
2
X
2
0.96
7
1
0.87
9
0.88
3
0.72
1
0.77
8
0.42
5
X
3
0.84
6
0.87
9
1
0.87
9
0.63
2
0.60
5
0.38
4
X
4
0.86
9
0.88
3
0.87
9
1
0.57
1
0.60
8
0.34
0
X
5
0.70
5
0.72
1
0.63
2
0.57
1
1
0.84
8
0.32
9
X
6
0.76
9
0.77
8
0.60
5
0.60
8
0.84
8
1
0.63
9
X
7
0.45
2
0.42
5
0.38
4
0.34
0
0.32
9
0.63
9
1
In Table 1, the seven preliminary evaluation
indicators of each dimension of profitability have
different focuses and are significantly different. The
correlation coefficient between X1 and X2 is 0.967,
indicating a highly positive correlation between
them. Similarly, the correlation coefficient between
X3 and X4 is 0.879, and the correlation coefficient
between X5 and X6 is 0.848. However, the
correlation between X7 and other dimensions is
relatively low, with a maximum of only 0.639. The
results of Bartlett's sphere test and KMO test are
illustrated in Table 2.
Table 2. Bartlett spherical and KMO inspection
results
Inspection category
Inspection
results
KMO sampling
0.785
Bartlett
sphericity
Approximate
chi-square
9402.524
Freedom
21
Significance
0
In Table 2, there are significant differences in the
profitability dimension of the seven primary
indicators. The KMO test results in a score of 0.782,
which is greater than 0.7 and suitable for FA.
Bartlett spherical test of the test results are divided
into three categories. The approximate chi-square is
9348.998, the degree of freedom is 21, and the
significance is 0. This indicates that the sampling
suitability of the sample is high, and the data fits
well under the spherical assumption. The Principal
Component Loadings Matrix and Score coefficient
matrix are shown in Figure 7.
WSEAS TRANSACTIONS on BUSINESS and ECONOMICS
DOI: 10.37394/23207.2024.21.14
Sujing Guo
E-ISSN: 2224-2899
159
Volume 21, 2024
X1 X2 X3 X4 X5 X6 X7
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
Component value
component matrix
Principal Component
Load Matrix
Principal Component
Score Matrix
Fig. 7: Principal component load matrix and score
coefficient
The principal components are represented by
1
X
,
2
X
,
3
X
,
4
X
,
5
X
,
6
X
. In terms of profitability,
the study selects the above five financial indicators
for the subsequent study. Then, following this
procedure, the preliminary evaluation indicators of
the other three dimensions are screened separately.
Finally, the indicators to be selected for each
dimension are combined and subjected to PCA
analysis. According to the criterion of a cumulative
variance contribution rate of 75% or more, this
paper has screened out the final 11 indicators from
the 27 preliminary indicators. The aim is to
construct a financial performance evaluation index
system for manufacturing enterprises.
4.2 Analysis of Financial Quality Results for
Manufacturing Companies
The 11 financial performance evaluation criteria
selected by the paper, are statically analyzed by
calculating the observations, the mean, the median,
the maximum, the minimum, and the standard
deviation. Among them, the statistics of the
indicators of the profitability dimension are shown
in Figure 8.
Return on
assets Net profit
margin of
total assets
Roe Return on
invested capital
Ratio of
Profits to
Cost
-100
-80
-60
-40
-20
0
20
40
60
80
100
Numerical value
Financial index
Average
Median
Maximum
Minimum
Standard
deviation
110
Fig. 8: Index statistics of profitability dimension
In Figure 8, the median of the overall return on
assets, the median return on assets, the total asset
net profit margin, and the cost expense profit margin
of manufacturing enterprises are 4.78%, 3.5%, and
7.12%, respectively. These median values are all
lower than the average of these three indicators,
indicating that some companies perform poorly in
terms of asset return, total asset net profit margin,
and cost expense profit margin. This situation has
led to the overall average of manufacturing
enterprises being pulled down. This also indicates
that some enterprises in the manufacturing industry
are facing problems such as low profitability, low
asset utilization efficiency, and poor cost
management. The statistics of the indicators of the
debt solvency dimension are shown in Figure 9.
Quick
ratio Asset
liability
ratio
-100
-80
-60
-40
-20
0
20
40
60
80
100
Average
Median Maximum
Minimum
Standard deviation
Financial index
Fig. 9: Index statistics of debt repayment ability
dimension
In Figure 9, the average levels of the quick ratio
and asset-liability ratio are 1.55 and 44.25%,
respectively. The asset-liability ratio is similar to its
corresponding reasonable levels of 1 and 50%. The
maximum value of the asset-liability ratio is 29.75.
The minimum value is 0.07, and the standard
deviation is 1.64. These results indicate that there
are significant differences in asset-liability ratio and
asset-liability ratio among listed companies in
China. Among them, the maximum value of the
asset-liability ratio is 98.21%, the minimum value is
1.43%, and the standard deviation is 18.41%. These
data show the differences in financial conditions and
changes in risk levels of listed companies. These
differences may be caused by factors such as
different industries, company sizes, and business
strategies. Therefore, when assessing financial
performance, it is necessary to take full account of
these differences and to analyze and make
judgments based on specific situations to develop
appropriate financial management and asset
WSEAS TRANSACTIONS on BUSINESS and ECONOMICS
DOI: 10.37394/23207.2024.21.14
Sujing Guo
E-ISSN: 2224-2899
160
Volume 21, 2024
allocation strategies for the company. The statistics
of the indicators of development ability and
operation ability dimension are shown in Figure 10.
Total Assets
Growth Rate Operating
revenue
growth rate
Current asset
growth rate
Total
Assets
Growth
Rate
(a)Statistical indicators of
development capacity (b)Statistics of operational capacity
indicators
-100
-80
-60
-40
-20
0
20
40
60
80
100
-100
-80
-60
-40
-20
0
20
40
60
80
100
Financial index Financial index
Average
Median Maximum
Minimum
Standard
deviation
110
Average
Median Maximum
Minimum
Standard
deviation
Numerical value
Numerical value
Fig. 10: Statistics of indicators for development and
operational capacity dimensions
In Figure 10, the mean value of the total assets
growth rate is 7.15 percent and the mean value of
the operating income growth rate is 21.06 percent,
with a standard deviation of 17.63 percent and 236.7
percent. There are significant differences between
the maximum and the minimum values of these two
indicators, which indicates that the average degree
of dispersion of these two indicators is high. The
mean value of the current assets turnover ratio and
total assets turnover ratio are 1.41 and 0.64
respectively. The median is 1.15 and 0.62
respectively. The mean value is greater than the
median, and the level of operational capacity of
most enterprises is high. The maximum value is
6.12, 3.41. The minimum value is 0.05, 0.04. The
standard deviation is 0.83, 0.34. Overall, the mean
value of each indicator is larger than the median in
the development capacity and operational capacity,
which indicates that the overall level of
development capacity and operational capacity of
enterprises in the manufacturing industry is
relatively strong. A comparison of the profitability
of digitised and non-digitized companies is carried
out for the period from 2014 to 2022.
In Figure 11, as a whole, the return on assets
(ROA) of digitally transformed firms averages
higher than the average of non-digitally transformed
firms across all years. Over the period 2014-2023,
the average return on equity (ROE) for digitally
transformed firms is generally higher than the
average ROE for non-digitally transformed firms, at
7.15 percent and 3.96 percent, respectively.
The ROE for non-digitally transformed
companies is 6.91 percent, while the ROE for non-
digitally transformed companies is 4.95 percent.
After 2020, digital transformation companies tend to
be more profitable than non-digital transformation
companies due to their expertise in digital
technology and their efficient cost management.
2022
202120202019
2018
2017201620152014
0
2
4
6
8
10
Year
Proportion
(a)Return on assets
2022
202120202019
2018
2017201620152014
0
2
4
6
8
10
Year
Proportion
(b)Net profit margin of total
assets
Digital
enterprise Non digital
enterprises Digital
enterprise Non digital
enterprises
2022
202120202019
2018
2017201620152014
0
2
4
6
8
10
Year
Proportion
(c)Roe
Digital
enterprise Non digital
enterprises
2022
202120202019
2018
2017201620152014
0
2
4
6
8
10
Year
Proportion
(d)Return on
invested capital
Digital
enterprise Non digital
enterprises
2022
202120202019
2018
2017201620152014
0
5
10
15
20
25
Year
Proportion
(d)Ratio of Profits to CostReturn on
assets
Digital
enterprise Non digital
enterprises
Fig. 11: Comparison of profitability between digital
and non-digital enterprises from 2014 to 2022
5 Conclusion
As a part of the real economy, manufacturing is
crucial in economic development. Every country in
the world has realized that the DT of enterprises has
become the inevitable choice of the times, and
combining the new generation of digital technology
with the traditional advantages of manufacturing
enterprises is a realistic need for the future progress
of enterprises. Under strong support for the digital
economy, traditional manufacturing enterprises have
started to utilize digital technology for business
transformation. In the digital economy, whether
digital transformation can bring new development
for enterprises, the changes in financial performance
before and after the transformation of manufacturing
enterprises are worth exploring. The study
employed the PCA method to select initial
indicators within the FPE index system for
manufacturing enterprises. After a thorough
screening, it constructed a financial performance
evaluation index system that suited the analysis of
digitally transformed enterprises. The paper
examined the differences in financial performance
before and after digital transformation and carried
out a comprehensive evaluation and comparative
analysis of financially transformed and non-digitally
transformed enterprises. The results indicated that
digitally transformed enterprises yielded a mean
return on net assets at a higher rate compared to
non-digitally transformed enterprises, specifically
7.15% and 3.96%, respectively. Furthermore,
WSEAS TRANSACTIONS on BUSINESS and ECONOMICS
DOI: 10.37394/23207.2024.21.14
Sujing Guo
E-ISSN: 2224-2899
161
Volume 21, 2024
digitally transformed enterprises appeared to boast
greater advantages and operational adaptability in
the digital arena when compared to their non-
digitally transformed counterparts. Digitally
transformed companies achieved superior financial
performance in comparison to their non-digitally
transformed counterparts. The stress capacity of
firms that have undergone digital transformation
was consistently higher than that of non-digitally
transformed firms across all years. However, the
study has some limitations. It is important to
approach all forms of evaluation methods
objectively and not excessively depend on them in
future research.
References:
[1] Li D, Chen Y, Miao J, Does ICT create a new
driving force for manufacturing? -Evidence
from Chinese manufacturing firms,
Telecommunications Policy, Vol.46, No.7,
2021, pp. 1029-1042.
[2] Jing S, Feng Y, Yan J, Path selection of lean
digitalization for traditional manufacturing
industry under heterogeneous competitive
position, Computers and Industrial
Engineering, Vol.161, No.2, 2021, pp. 1631-
1648.
[3] Jafari S V, Garcia A, Candelo E, Couturier J,
Exploring the impact of digital transformation
on technology entrepreneurship and
technological market expansion: the role of
technology readiness, exploration and
exploitation, Journal of Business Research,
Vol.124, No.1, 2021, pp. 100-111.
[4] Long G, Li C, Li S, Xu T, Nonlinear
characteristics of the effect of manufacturing
servitization on consumer business
performance, Mathematical Problems in
Engineering, Vol.2021, NO.49, 2021, pp. 41-
52.
[5] Li J P O, Liu H, Ting D S J, Jeon S, Chan R V
P, Kim J E, Digital technology, tele-medicine
and artificial intelligence in ophthalmology: a
global perspective, Progress in Retinal and
Eye Research, Vol.82, No.3, 2020, pp. 25-86.
[6] Lim M K, Xiong W, Wang C, Cloud
manufacturing architecture: a critical analysis
of its development, characteristics and future
agenda to support its adoption, Industrial
Management and Data Systems, Vol.121,
No.10, 2021, pp. 2143-2180.
[7] Li J, Saide S, Ismail M N, Indrajitr R E,
Exploring IT/IS proactive and knowledge
transfer on enterprise digital business
transformation (EDBT): a technology-
knowledge perspective, Journal of Enterprise
Information Management, Vol.35, No.2,
2021, pp. 597-616.
[8] Chen W, Zhang L, Jiang P, Meng F, Sun Q,
Can digital transformation improve the
information environment of the capital
market? Evidence from the analysts'
prediction behavior, Accounting and Finance,
Vol.62, No.2, 2021, pp. 2543-2578.
[9] Anthony Jnr, B., Petersen S A, Helfert M,
Guo H, Digital transformation with enterprise
architecture for smarter cities: a qualitative
research approach, Digital Policy, Regulation
and Governance, Vol.23, no.4, 2021, pp. 355-
376.
[10] Sia S K, Weill P, Zhang N, Designing a
future-ready enterprise: the digital
transformation of DBS bank, California
Management Review, Vol.63, No.3, 2021, pp.
35-57.
[11] Dilek C K, Babak A, Understanding the role
of employees in digital transformation:
conceptualisation of digital literacy of
employees as a multi-dimensional
organizational affordance, Journal of
Enterprise Information Management, Vol.34,
No.6, 2021, pp. 1649-1672.
[12] Laghi E, Marcantonio M D, Cillo V, Paoloni
N, The relational side of intellectual capital:
an empirical study on brand value evaluation
and financial performance, Journal of
Intellectual Capital, Vol.23, No.3, 2020, pp.
479-515.
[13] Cremers K J M, Fulkerson J A, Riley T B,
Benchmark discrepancies and mutual fund
performance evaluation, Journal of Financial
and Quantitative Analysis, Vol.57, No.2,
2021, pp. 543-571.
[14] Jahangir M, Mokhtari R, Mousavi S A,
Performance evaluation and financial analysis
of applying hybrid renewable systems in
cooling unit of data centres-a case study,
Sustainable Energy Technologies and
Assessments, Vol.46, NO.8, 2021, pp. 32-46.
[15] Meng Q, Zhou Y, Enterprise economic
performance evaluation based on 5G network
and embedded processing system,
Microprocessors and Microsystems, Vol.80,
No.2, 2021, pp. 61-65.
[16] Mousavi M M, Lin J, The application of
PROMETHEE multi-criteria decision aid in
financial decision making: case of distress
prediction models evaluation, Expert Systems
with Applications, Vol.159, No.11, 2020, pp.
WSEAS TRANSACTIONS on BUSINESS and ECONOMICS
DOI: 10.37394/23207.2024.21.14
Sujing Guo
E-ISSN: 2224-2899
162
Volume 21, 2024
3438-3458.
[17] Wang H, Feng J, Zhang H, Li X, The effect of
digital transformation strategy on
performance: the moderating role of cognitive
conflict, International Journal of Conflict
Management, Vol.31, No.3, 2020, pp. 441-
462.
[18] Zhu L, Li M, Metawa N, Financial risk
evaluation z-score model for intelligent IoT-
based enterprises, Information Processing and
Management, Vol.58, No.6, 2021, pp. 102-
111.
[19] Birskyte L, Mingelaite D, Government
financial support: does it improve the
performance of small and medium-sized
enterprises in Lithuania? International
Journal of Entrepreneurship and Small
Business, Vol.43, No.1, 2021, pp. 142-163.
[20] Choudhuri S, Adeniye S, Sen A, Distribution
alignment using complement entropy
objective and adaptive consensus-based label
refinement for partial domain adaptation,
Artificial Intelligence and Applications, Vol.1,
No.1, 2023, pp. 43-51.
Contribution of Individual Authors to the
Creation of a Scientific Article (Ghostwriting
Policy)
The author contributed in the present research, at all
stages from the formulation of the problem to the
final findings and solution.
Sources of Funding for Research Presented in a
Scientific Article or Scientific Article Itself
No funding was received for conducting this study.
Conflict of Interest
The authors have no conflict of interest to declare.
Creative Commons Attribution License 4.0
(Attribution 4.0 International, CC BY 4.0)
This article is published under the terms of the
Creative Commons Attribution License 4.0
https://creativecommons.org/licenses/by/4.0/deed.en
_US
WSEAS TRANSACTIONS on BUSINESS and ECONOMICS
DOI: 10.37394/23207.2024.21.14
Sujing Guo
E-ISSN: 2224-2899
163
Volume 21, 2024