The Asian Journal of Shipping and Logistics xxx (xxxx) xxx

Contents lists available at ScienceDirect

The Asian Journal of Shipping and Logistics

journal homepage: [Link]/locate/ajsl

Logistics scale as trade catalyst: Principal component analysis of

China-Korea bilateral trade drivers
Yue Zhang a,*, Nianjie Shang b, Yanting Liu c , Chao Zhang a , Lu Jin a
a
Department of International Trade and Economics, Silla University, Busan 46958, South Korea
b
Department of Global Food Service, Management, Woosuk University, Wanju 55338, South Korea
c
Department of International Area, Silla University, Busan 46958, South Korea

A R T I C L E I N F O A B S T R A C T

Keywords: This study conducts an empirical analysis of the driving factors and development trends of China-Korea bilateral
Bilateral trade trade from the perspective of logistics, using IBM SPSS Statistics 27 for data processing and modeling. Based on
China-South Korea trade annual panel data from 2013 to 2022, a multidimensional dataset of 18 core economic indicators was con­
Logistics sector
structed, encompassing variables such as import and export volumes, logistics scale, enterprise expenditures,
Multiple regression analysis
Logistics enterprises
labor structure, exchange rates, and investment. To reduce dimensionality and address multicollinearity, a
Employment in logistics correlation analysis was first conducted to identify 9 key variables with strong explanatory power. Principal
Logistics sales revenue Operational expenses Component Analysis was then applied, extracting five principal components that effectively captured the un­
Container handling derlying structure of the dataset. These components were used as independent variables in a principal component
regression model to predict the China-Korea bilateral trade volume. Among them, the component representing
Korea’s logistics industry scale and transportation efficiency (PC1) had the most significant positive impact on
trade, highlighting the critical role of logistics in enhancing trade intensity and sustaining bilateral growth. The
regression model exhibited strong goodness-of-fit and low prediction errors, through structural modeling and
trend-based forecasting, this study confirms the foundational role of Korea’s logistics system in supporting
bilateral trade and provides quantitative insights and policy-oriented recommendations for future China-Korea
economic cooperation.

1. Introduction South Korea with greater access to international trade opportunities.

Among these, Chinese e-commerce platforms have gained a dominant
China and South Korea share geographic proximity, which greatly share in Korea’s direct purchase market, offering competitive pricing
facilitates bilateral trade. According to data released by South Korea’s and lower shipping costs (Shikur, 2022).
Ministry of Trade, Industry and Energy, the country’s total trade volume According to Statistics Korea, the total value of online shopping
in 2023 reached USD 1279.53 billion. Of this, exports to China reached KRW 137 trillion in 2019 and nearly doubled to KRW 210
amounted to USD 124.835 billion, while imports from China totaled trillion by 2022. The cross-border direct purchase market also saw
USD 142.849 billion—firmly establishing China as South Korea’s largest substantial growth, exceeding KRW 6 trillion for the first time.
trading partner and underscoring the high level of trade interdepen­ Regionally, imports from China accounted for KRW 3.2873 trillion in
dence between the two economies (Ministry of Trade, Industry and 2022, representing a 121.2 % year-on-year increase. During this period,
Energy, 2023). China surpassed the United States to become South Korea’s leading
As a nation heavily reliant on foreign trade, South Korea’s logistics source of cross-border purchases (Chen et al., 2018; Zhang, 2020).
sector is closely tied to the volume of international orders. With the Chinese e-commerce platforms such as AliExpress and Temu have
increasing sophistication of global division of labor, logistics and trade rapidly expanded into the Korean market, capitalizing on their pricing
have begun to evolve into distinct yet interdependent industries. The advantage. As a result, South Korea’s imports of cross-border e-com­
rise of cross-border e-commerce has further broken-down geographic merce goods from China grew sharply in 2022, generating a surge in
barriers, enabling faster and more efficient transactions and providing logistics demand. This expansion has significantly contributed to the

* Corresponding author.
E-mail address: isee008@[Link] (Y. Zhang).

[Link]
Received 8 October 2024; Received in revised form 19 August 2025; Accepted 1 September 2025
2092-5212/© 2025 The Authors. Published by Elsevier B.V. on behalf of The Korean Association of Shipping and Logistics, Inc. This is an open access article under
the CC BY-NC-ND license ([Link]

Please cite this article as: Yue Zhang et al., The Asian Journal of Shipping and Logistics, [Link]
Y. Zhang et al. The Asian Journal of Shipping and Logistics xxx (xxxx) xxx

development of logistics not only within South Korea but also across 2.1. Key determinants of logistics-driven trade
neighboring countries, strengthening South Korea’s competitiveness in
the international logistics arena. Drawing on extant literature, the following variables are identified as
Logistics scale efficiency is a critical driver of trade growth, primarily empirically substantiated drivers of trade-logistics interdependence. The
by reducing transportation costs and enhancing delivery speed bilateral trade volume between China and South Korea serves as a direct
(Zablotska & Mammadov, 2017). Transportation costs, a major measure of trade intensity and reflects the level of economic interde­
component of international logistics expenses, directly influence pendence between the two nations. Studies have shown a strong cor­
corporate profitability and trade volume. Rising transport costs increase relation between logistics efficiency and trade facilitation (Shikur,
export prices, eroding competitiveness in global markets, whereas cost 2022). South Korea’s Gross Domestic Product (GDP) indicates its eco­
reductions enhance trade performance. Inefficient labor directly leads to nomic capacity and demand, both of which directly influence trade
a slowdown in the loading and unloading speed of cargo at ports, flows and logistics needs, in line with previous findings on the rela­
reducing the overall throughput capacity of the ports (Bottalico, 2022). tionship between economic scale and trade volume (Katrakylidis &
However, international trade is also shaped by other factors, including Madas, 2019; Ma et al., 2020).
policy shifts, political dynamics, and macroeconomic conditions in Furthermore, the number of employees and enterprises in the lo­
trading nations (Liang & Liu, 2020). gistics sector reflects the industry’s scale and human capital—factors
As key nodes in global trade, ports play a pivotal role, where fluc­ crucial to improving operational efficiency. This aligns with existing
tuations in maritime freight costs directly impact import/export oper­ research that emphasizes the role of labor capacity and organizational
ations and broader trade patterns. Port efficiency and trade exhibit density in enhancing logistics performance (Gacuru and Kabare, 2015;
complementarity, generating synergistic effects that amplify regional Song and Lee, 2022). Logistics costs and sales figures quantify the sec­
economic and social welfare (Kojo Ayesu et al., 2022). Modern port tor’s economic input and output, and are directly linked to trade
automation technologies including automated gantry cranes, smart competitiveness and cost-effectiveness. As shown in prior studies, lo­
container tracking systems, and AI-driven scheduling algorithms - have gistics spending and revenue flows significantly influence supply chain
demonstrated effectiveness in enhancing this synergy. For instance, feasibility and overall trade performance (Celebi, 2019; Mahpour et al.,
Chile’s automated terminal operations reduced vessel turnaround time 2021).
while increasing annual throughput capacity (Duran et al., 2024). Container port throughput serves as an indicator of infrastructure
Nevertheless, high port throughput alone does not guarantee trade capacity and connectivity—key determinants of trade facilitation. A
expansion; inefficiencies in logistics operations—such as delays in cargo well-developed transport infrastructure underpins efficient logistics
handling, customs clearance bottlenecks, or poor warehousing man­ operations (Mahpour et al., 2021). Scholars have specifically explored
agement—can hinder the full realization of trade potential (Celebi, the impact of China’s inland port development on China-South Korea
2017). When such inefficiencies persist, cargo flow is disrupted, trade flows, suggesting that improvements in China’s logistics infra­
diminishing the positive correlation between port capacity and trade structure will further ease bilateral trade and support its continued
growth, even in highly automated facilities. growth.
Efficient international logistics plays a vital role in promoting global Trade advancement also supports logistics. One empirical study used
trade. With the advancement of economic globalization, logisti­ the Logistics Performance Index (LPI) to analyze how export promotion
cs—especially those driven by e-commerce—has become a critical policies affected trade among Korea, China, and Japan. The findings
enabler of cross-border commerce, helping bridge regional resource highlighted the crucial role of trade facilitation in improving logistics
disparities and even reshape international trade dynamics (Celebi, 2019; performance. Strengthened logistics partnerships can improve regional
Gacuru and Kabare, 2015; Song and Lee, 2022). In the context of connectivity and promote economic integration.
China-South Korea trade relations, efficient logistics systems are essen­ Additionally, cross-border logistics cooperation enhances mutual
tial for sustaining and expanding bilateral trade. This study adopts a engagement and competitiveness. The gravity model has been widely
regression-based empirical approach to identify key factors influencing used to assess the economic feasibility of logistics integration among
trade-related logistics between China and South Korea, offering flexible China, Korea, and Japan (Kim et al., 2021). By quantifying the economic
and sustainable logistics strategies for businesses and policymakers alike benefits of such cooperation, this model helps evaluate the practicality
(Ma et al., 2020). and strategic value of multilateral logistics collaboration.
The structure of this article is as follows: 2 provides a brief review of In summary, prior research has offered valuable insights into China’s
the literature on China-South Korea bilateral trade and logistics. 3 in­ influence on Korea’s logistics development from the perspectives of
troduces the data sources, the principal component analysis (PCA) trade cooperation, infrastructure improvement, export promotion, and
methodology, and the regression model. 4 presents the empirical results, logistics efficiency. The indicators selected for this study include 18
highlighting the decisive role of logistics in shaping bilateral trade. 5 indicators, including China-South Korea bilateral trade volume, South
discusses the limitations of the model and proposes relevant policy Korea’s GDP, number of people employed in the logistics industry,
recommendations. Finally, the paper concludes with a summary of the number of logistics companies, logistics costs, logistics sales, and
findings and outlines potential directions for future research. While this container port throughput (Table 1. All variables affecting China-Korea
approach offers robust insights, three inherent limitations warrant bilateral trade). These indicators are grounded in well-established
upfront acknowledgment: (1) temporal aggregation may obscure short- literature on logistics performance and international trade dynamics.
term shocks, (2) Limited policy factors lead to inaccurate and inaccurate Understanding these factors is essential for improving logistics effi­
results. ciency, reducing trade costs, and fostering regional economic growth.
South Korea’s well-developed logistics infrastructure—resulting from
2. Literature review years of investment and innovation—has secured its competitive posi­
tion in the global logistics landscape.
This section synthesizes prior research through two interconnected
lenses: 2.1 empirically validated variables shaping logistics-trade dy­ 2.2. Mathematical models
namics, and 2.2 methodological approaches for quantifying these re­
lationships. This structure directly supports modeling strategy by The relationship between trade development and logistics perfor­
grounding variable selection in established theory and justifying the mance has long been a focal point in international economics and supply
PCA-regression framework. chain management research. Numerous studies have confirmed that
efficient logistics systems are key enablers of international trade growth,

2
Y. Zhang et al. The Asian Journal of Shipping and Logistics xxx (xxxx) xxx

Table 1 3. Status of logistics development in Korea

All variables affecting China-Korea bilateral trade.
No. Variable Supporting Literature South Korea’s rapid development in international trade is closely
linked to the expansion and modernization of its logistics sector. Both
1 China-Korea bilateral trade（USD Guo, X., Li, J., & Huang, S. (2023).
100 million） domestic and international scholars generally agree that modern logis­
2 Korea’s GDP（USD 100 million） Kim, D., Cho, G., & Park, Y. (2021) tics serves as a catalyst for economic development. The widely recog­
3 China’s GDP（USD 100 million） Kim, D., Cho, G., & Park, Y. (2021) nized "logistics-driven growth theory" argues that the advancement of
4 China’s imports to Kore（USD Wang, Y., & Xin, L. (2020). logistics deepens the specialization of labor, reduces the cost of import
100 million）
5 China-Korea bilateral trade Wang, Y., & Xin, L. (2020).
and export operations, and ultimately drives the high-quality growth of
6 Logistics sales（USD 100 Song, D., & Lee, S. (2022) foreign trade. In recent years, South Korea has made notable progress in
million） building a well-structured and highly efficient logistics system. By
7 Value added of logistics (USD 100 Song, D., & Lee, S. (2022) introducing advanced technologies and learning from global best prac­
million)
tices, Korea has maintained a high standard in its logistics environment
8 Number of Logistics Enterprises Wang, F., & Liu, Q
9 Number of employees (10,000) Zhang, X. (2020) on the international stage.
10 Total logistics costs (USD 100 Zhang, X. (2020) Today, the country operates a comprehensive logistics network
million) supported by three major systems: seaport, airport, and overland
11 Logistics material costs (USD 100 Shen, Y., Liu, J., & Tian, W. (2022). transportation. South Korea is home to two world-class logistics
million)
12 Logistics fuel costs (USD 100 Shen, Y., Liu, J., & Tian, W. (2022).
hubs—Busan Port, the second-largest transshipment port in the world,
million) and Incheon International Airport, the third-largest cargo airport glob­
13 Logistics Employee Costs (USD Shen, Y., Liu, J., & Tian, W. (2022). ally. These infrastructure advantages enable faster and more cost-
100 million) effective import and export processing, providing favorable conditions
14 Cargo throughput at Korean ports Lee, S., & Song, D. (2017)
for companies engaged in trade activities in South Korea.
(USD 10,000 T)
15 Korea’s investment in China (USD Lee, S., & Song, D. (2017) The Logistics Performance Index (LPI), published biennially by the
100 million) World Bank, is a widely accepted indicator of a country’s logistics
16 Chinese investment in Korea (USD Kim, H., & Lee, Y. (2016) competitiveness. It evaluates six core components: customs perfor­
100 million) mance, infrastructure quality, international shipment ease, logistics
17 China-Korea Exchange Rate Kim, H., & Lee, Y. (2016)
services, tracking and tracing, and timeliness, across approximately 160
（USD 100 million）
18 Cargo throughput at Chinese ports Shamohammadi, M., Kwon, O. K., & countries. According to the 2023 LPI report, South Korea ranked 17th
(10,000 T) Lee, S. Y. (2025) globally in overall logistics performance. Notably, its performance in
customs clearance (ranked 7th) and logistics infrastructure (ranked 9th)
placed the country within the global top 10. While Korea’s LPI ranking
especially in export-oriented economies like South Korea (Celebi, 2019).
had fallen in previous years—21st in 2014, 24th in 2016, and 25th in
Logistics infrastructure, port capacity, and human capital are among the
2018—and was excluded during the COVID-19 period, the 2023 result
core elements that influence the efficiency of cross-border trade.
marked a significant recovery with an 8-position rise, securing a place
South Korea, as a highly trade-dependent economy, has undergone
among the world’s top 20 for the first time.
significant transformation in its logistics industry to support its role in
Data on the number of logistics companies and logistics employees
global supply chains. Song and Lee emphasized the role of labor and
further illustrates the sector’s rapid expansion. A surge in newly regis­
organizational density in enhancing logistics performance, while Mah­
tered firms points to a dynamic and competitive market, attracting
pour et al. (2021) discussed how logistics costs directly affect trade
numerous new entrants. At the same time, the continued rise in
competitiveness in East Asia. Meanwhile, China’s increasing investment
employment indicates growing labor demand within the sector.
in inland ports and digital trade platforms has also reshaped the regional
Although the growth rate of employee numbers has not matched that of
logistics landscape, strengthening its trade connectivity with Korea
firms, the overall increase remains significant. This discrepancy may be
(Zhang, 2020).
due to automation and technological advancement, allowing individual
Several researchers have utilized the gravity model of trade to
firms to operate more efficiently with fewer employees. Nonetheless,
explain bilateral trade flows, showing that ecoomic size and logistics-
total employment continues to rise, reflecting the sector’s sustained
related frictions (such as distance, infrastructure, and customs proced­
demand for human resources. The near doubling of logistics companies
ures) are significant determinants of trade volumes (Katrakylidis and
underscores a sharp increase in market activity, while the steady growth
Madas, 2019; Ma et al., 2020). Building on this, Liu applied a modified
in employment highlights the industry’s evolving balance between labor
gravity model to examine the economic rationale of trilateral logistics
cooperation among China, Korea, and Japan, concluding that integrated
logistics policies can significantly boost regional trade.
Principal Component Analysis is often used to handle multi­
collinearity in high-dimensional trade and logistics data. Gacuru and
Kabare (2015) highlighted the importance of combining macroeco­
nomic indicators with logistics variables to better explain trade
behavior. Recent studies by Chen and Hu (2016) further employed
composite logistics indices and regression analysis to assess how logis­
tics performance metrics such as the Logistics Performance Index (LPI)
impact bilateral export potential.
Overall, the existing literature confirms the bidirectional relation­
ship between logistics development and trade expansion, emphasizing
the need for robust empirical modeling. This study builds upon these
insights by employing PCA-based regression models to investigate the
key logistics factors influencing China-Korea bilateral trade, thereby
providing new quantitative evidence and policy implications.
Fig. 1. Number of logistics companies and employees in Korea. Source: Min­
istry of Commerce of China.

3
Y. Zhang et al. The Asian Journal of Shipping and Logistics xxx (xxxx) xxx

and technology (Fig. 1 Number of logistics companies and employees in

Korea).
Container throughput is a key indicator for assessing freight flow and
transportation capacity, directly reflecting the overall efficiency of a
country’s logistics system. From 2017–2021, South Korea’s port cargo
volume exhibited a consistent upward trend, increasing from 2.742
billion tons to 2.979 billion tons. However, in 2022, this figure slightly
declined to 2.85 billion tons—the lowest level in five years. This
downturn may be attributed to external disruptions such as typhoons
and other operational challenges that affected port activities during that
year (Fig. 2 Container volume in Korea, 2013–2022).
Beyond port throughput, a comprehensive review of macro logistics
indicators from 2013 to 2022 reveals a steady expansion of Korea’s lo­
gistics industry. The number of logistics enterprises increased from
359,587 in 2013–422,344 in 2022, reflecting growing market partici­
Fig. 3. Number of logistics companies and employees in Korea. （Unit：USD
pation and a competitive business environment. Likewise, employment
billion）. Source: Ministry of Commerce of China.
in the logistics sector rose from approximately 1.09 million people to
over 817,000—despite some fluctuation likely due to automation and
structural transformation. While the number of enterprises nearly Table 2
doubled, employment grew at a slower pace, which may suggest Descriptive Statistics of Core Variables for Logistics-Driven Trade.
increased operational efficiency driven by technological advancement. Variable Name Unit Description r
In terms of cost structure, total logistics expenses increased signifi­
Sales Revenue USD 100 Reflects output level of 0.911
cantly from USD 930.08 billion in 2013 to nearly USD 1392.34 billion in million logistics services
2022. Fuel and labor costs also rose, indicating continued input into Operating Expenses USD 100 Main financial outflows in 0.790
transportation infrastructure and human capital. During the same million logistics
period, the volume of international freight handled by Korea rose from Material Costs USD 100 Physical input into logistics 0.788
million operations
approximately 1.13 million tons to over 1.31 million tons, further
Fuel Costs USD 100 Energy expenditures 0.761
demonstrating the country’s expanding role in global logistics chains million
(Fig. 3. Number of logistics companies and employees in Korea). Labor Costs USD 100 Human resource expenditures 0.708
The growth of container throughput and logistics investments aligns million
with South Korea’s positioning as a regional logistics hub, supported by Number of Logistics Firms Industry scale and market 0.679
Enterprises saturation
world-class infrastructure such as Busan Port and Incheon International
Number of Logistics 10,000 Labor force input 0.558
Airport. However, the slight downturn in 2022 highlights the vulnera­ Employees people
bility of even advanced systems to environmental and geopolitical risks, Port Throughput 100 million Measures infrastructure 0.543
reinforcing the need for resilience-oriented logistics strategies. (Korean Ports) tons capacity
Volume of International Tons Reflects cross-border logistics 0.515
Cargo activity
4. Principal component analysis

In constructing the indicator system, attention is paid to covering (see Table 2 Descriptive Statistics of Core Variables for Logistics-Driven
multiple dimensions of logistics economic characteristics, including Trade). For highly correlated variables, in accordance with the principle
cost, human resources, infrastructure, import-export, and industry scale, of statistical variable simplification, only the indicator with stronger
to ensure a comprehensive reflection of such characteristics while representativeness is retained to eliminate the interference of multi­
avoiding dimensional redundancy. Using SPSS Statistics 27 and based on collinearity. Meanwhile, indicators with no significant association or a
the results of Pearson correlation coefficient analysis, indicators that negative correlation with sales are strictly excluded to avoid the risk of
were significantly positively correlated with sales were prioritized, as insufficient explanatory power or misleading conclusions. To guide this
these variables exhibit strong explanatory power for sales fluctuations process intuitively, Fig. 4 systematically illustrates the indicator
screening process based on principal component analysis (PCA).
Through the above screening strategies, an indicator system that com­
bines scientificity and effectiveness is constructed, providing a reliable
analytical framework for revealing the internal relationship between the
logistics economy and sales.
Significant multicollinearity was detected among the initial 18 var­
iables, with six pairwise correlations exceeding 0.7, indicating sub­
stantial conceptual overlap. To address this, Principal Component
Analysis was implemented following standardization and eigenvalue
decomposition of the covariance matrix. The analysis yielded five
orthogonal components (all inter-component correlations < 0.3),
collectively explaining 93.7 % of the total variance. This approach ad­
heres to established dimensionality reduction protocols for correlated
economic indicators (Jolliffe & Cadima, 2016), effectively eliminating
multicollinearity while preserving explanatory power (Table 3. Break­
down of explained variance):
The component loadings indicate the correlation between the orig­
inal variables and the principal components. A summary of the loadings
Fig. 2. Container volume in Korea, 2013–2022. (Unit:million TEUs). Source: is shown in the table below (Table 4. Summary of the loadings):
Ministry of Commerce of China. This loadings table is crucial for interpreting the meaning of

4
Y. Zhang et al. The Asian Journal of Shipping and Logistics xxx (xxxx) xxx

Fig. 4. PCA-Based Indicator Selection Workflow.

efficiency. It includes indicators like international freight volume,

Table 3 intermodal transport coverage, and delivery timeliness for international
Breakdown of explained variance.
shipments, reflecting competitiveness in global supply chain networks.
Principal Component Explained Variance (%) Cumulative Variance (%) These five principal components serve as the independent variables in
PC1 51.3 51.3 the principal component regression model used in 5. This trans­
PC2 17.6 68.9 formation reduces dimensionality and avoids multicollinearity,
PC3 11.8 80.7 improving the stability and predictive power of the trade model (Kim &
PC4 7.3 88.0
PC5 5.7 93.7
Lee, 2016). The subsequent section will present regression results link­
ing these components to the bilateral trade volume between China and
South Korea.

Table 4
Summary of the loadings. 5. Regression analysis based on principal components

Variable PC1 PC2 PC3 PC4 PC5

To analyze the influence of logistics performance on bilateral trade, a
Sales Revenue 0.45 0.03 − 0.11 0.02 − 0.02 multiple linear regression model is constructed using the five extracted
Operating Expenses 0.43 0.08 0.04 0.01 0.15
principal components (PC1 to PC5) as independent variables. The
Material Costs 0.41 0.14 − 0.17 − 0.05 0.08
Labor Costs 0.38 0.07 0.15 0.13 − 0.02 dependent variable is the China-Korea bilateral trade volume (in USD
Number of Logistics Enterprises 0.40 − 0.08 0.13 − 0.25 − 0.02 100 million). The model is specified as follows:
Number of Logistics Employees 0.37 0.10 0.14 − 0.18 − 0.03
Port Throughput 0.35 − 0.19 0.11 0.68 0.32 Trade = β₀ + β₁PC1t + β₂PC2t + β₃PC3t + β₄PC4t + β₅PC5t + εt (1)
Volume of International Cargo 0.34 0.06 0.49 0.12 0.68
Where denotes the bilateral trade volume in year, represents the
score of the t principal component, and is the error term.
principal components. Each loading value indicates the strength and From the Table 5, it can observe that PC1’s coefficient is 138.00 and
direction of the relationship between an original variable and a specific is highly significant at the 0.01 level, indicating a significant positive
principal component. A larger absolute value means the variable has a impact of PC1 on the dependent variable. Similarly, PC3’s coefficient is
stronger influence on that component. Loadings ≥ |0.3| indicate − 139.82 and is also highly significant at the 0.01 level, suggesting a
meaningful component-variable relationships. A more detailed inter­ significant negative impact of PC3 on the dependent variable. PC5’s
pretation of the bracketed meanings is provided below to enhance coefficient is 52.04 and is significant at the 0.05 level, showing a positive
clarity and contextual depth: effect on the dependent variable. In contrast, the coefficients for PC2 and
PC1 (Logistics Scale & Cost Capacity): This principal component PC4 are not significant.
encapsulates the overall logistics scale and cost-bearing capability of PC1 shows a significantly positive effect on trade volume, indicating
firms. It primarily captures variables such as enterprise size, sales per­ that larger, more resource-intensive logistics systems promote bilateral
formance, and cost-related indicators. which collectively reflect the trade.PC3, which reflects variance in employee compensation and in­
operational scale of logistics enterprises and their capacity to manage ternational cargo structure, shows a significant negative effect, sug­
costs while sustaining business activities. gesting potential inefficiencies or mismatches.PC5 associated with
PC2 (Minor Cost Fluctuation Component): This component exhibits transport efficiency and infrastructure capacity, has a significant posi­
weak correlations with the original variables, suggesting limited tive effect. The insignificance of PC2 and PC4 reveals nuanced insights:
explanatory power over financial dynamics. Though nominally linked to PC2’s weak impact suggests short-term cost fluctuations (e.g., fuel price
fluctuations in material costs and resource allocation patterns, its low swings) are absorbed by large Korean logistics firms through economies
statistical significance precludes robust interpretation as a driver of of scale, minimizing trade disruption (Shamohammadi et al., 2025).
structural volatility. Consequently, PC2 is retained primarily for meth­ PC4’s non-significance may reflect saturation effects at major ports like
odological completeness in variance accounting. Busan—where throughput expansion beyond current levels yields
PC3 (Labor-Related Inefficiencies): This principal component high­ diminishing returns for trade growth without complementary infra­
lights inefficiencies originating from labor-related factors, such as low structure upgrades (S.H. Kang et al., 2021). These results demonstrate
productivity, workforce skill gaps, or suboptimal labor allocation. It may that specific dimensions of logistics—especially scale and international
incorporate variables like employee turnover rates, labor cost-to-output transport capacity—play a vital role in shaping trade outcomes between
ratios, or delays caused by human resource constraints, thereby
reflecting challenges in human capital management that impact opera­ Table 5
tional efficiency. Regression results.
PC4 (Port Throughput): Dominated by port cargo handling capacity,
Variable Coefficient Standard Error t-Statistic Significance
this component directly reflects the infrastructure capability of logistics
Intercept () 2846.36 71.22 39.94 ***
hubs. Variables such as annual port throughput (in tonnage or container
PC1 138.00 22.54 6.12 ***
volume), berth utilization rates, and cargo clearance efficiency load PC2 − 4.21 19.88 − 0.21 -
heavily on PC4, serving as key metrics for evaluating the operational PC3 − 139.82 27.67 − 5.05 ***
capacity and connectivity of maritime or inland port systems. PC4 − 4.37 13.13 − 0.33 -
PC5 (International Transport Capacity): This component quantifies a PC5 52.04 16.88 3.08 **

region’s or enterprise’s ability to engage in international logistics op­ Note: *p < 0.1
erations, focusing on cross-border cargo volume and transportation **p < 0.05
***p < 0.01

5
Y. Zhang et al. The Asian Journal of Shipping and Logistics xxx (xxxx) xxx

China and South Korea (Lim, C., & Lee, S, 2021). Despite these discrepancies, the model remains a reliable tool for
Based on the regression results using the five principal components, interpreting logistics-driven trade dynamics. Notably, the predicted
the estimated model for bilateral trade is: values from 2020 to 2030 exhibit a growth pattern consistent with
historical trends, validating the effectiveness of the principal component
Trade= 2846.36+ 138.00PC1− 4.21PC2− 139.82PC3− 4.37PC4+ 52.04­ regression model in economic forecasting involving multiple correlated
PC5 (2)
variables.
To forecast the trends in China-Korea bilateral trade volume, this
study employed the Principal Component Regression method. Using the 6. Discussion and policy implications
actual trade volume as the dependent variable, a regression model was
constructed, yielding the following predictive formula. The following 6.1. Model limitations
Fig. 5 compares the predicted bilateral trade values with the actual
observed data from 2013 to 2022, using the regression model based on Although the PCA-based regression model effectively reduces
principal components. dimensionality and addresses multicollinearity, it also presents several
In this equation, the constant term 2846.36 represents the baseline limitations. First, the interpretation of principal components is abstract,
trade volume, PC1 through PC5 are the standardized principal compo­ which may obscure the specific role of individual indicators. Second, the
nent scores, and the coefficients indicate the marginal impact of each dataset is limited to annual data from 2013 to 2022, which may not fully
component on trade volume. As illustrated in the figure, the orange line capture short-term fluctuations or policy shocks. Additionally, external
represents the actual trade volume between China and Korea, while the factors such as global trade tensions and pandemic-related disruptions
grey line indicates the predicted values generated from the model. The are not explicitly modeled, though they may influence both logistics and
dotted orange line shows the linear trend of the actual data. trade performance.
n
1∑ 1 ∑ 6.2. Implications
MAE = yi| =
|yi − ̂ |Actual − Predicted| ≈ 177.5 (3)
n n=1 11
The regression results highlight several logistics-related priorities for
√̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅
1∑ n enhancing bilateral trade, with critical insights integrated from the
RMSE = (yi − ̂ y i )2 ≈ 222.1 (4) challenges of low circulation speed, insufficient logistics infrastructure
n n=1
(particularly in cold chain), lack of integrated information platforms,
⃒
100% ∑⃒⃒yi − ̂
⃒
y i ⃒⃒ scarce integrated services and strategic alliances, and frequent trade
MAPE = ≈ 6.56% (5) frictions in China-South Korea cross-border logistics:
n ⃒ yi ⃒
Investment in large-scale logistics infrastructure: Given the
R2 = 0.963 (6) strong positive effect of PC1, policies must prioritize upgrading cross-
border cold chain facilities and building smart logistics hubs to
To evaluate the prediction accuracy of the principal component address infrastructure gaps. For example, co-developing intelligent cold
regression model, four common error metrics were employed: Mean storage centers and dedicated cross-border cold transport lines in ports
Absolute Error, Root Mean Squared Error, Mean Absolute Percentage like Qingdao and Incheon can reduce fresh product spoilage rates while
Error, and the coefficient of determination (R2). The results indicate that enhancing operational scale. This aligns with the need to improve
the model achieved an MAE of 177.5, an RMSE of 222.1, and a MAPE of physical logistics efficiency amid low circulation speeds, as supported by
approximately 6.56 %, all of which point to relatively small deviations research showing infrastructure upgrades directly boost trade volumes
between actual and predicted values. The high R2 value of 0.963 further (Lim, C., & Lee, S., 2021).
suggests that over 96 % of the variance in the observed trade volume can Enhancing international cargo handling and digital connectiv­
be explained by the model, confirming its strong predictive performance ity: The significance of PC5 underscores the urgency of integrating
and reliability for economic forecasting purposes. cross-border logistics information platforms to solve data fragmentation.
The figure demonstrates that the model reasonably captures the Upgrading port logistics must involve not only hardware improvements
overall fluctuations and growth trends in bilateral trade. It aligns closely (e.g., expanding international shipping routes) but also establishing
with observed values in years such as 2013, 2017, and 2022, indicating shared digital systems—such as a China-South Korea logistics data
the strong explanatory power of the selected principal components. hub—that unify e-waybill standards and real-time tracking. This not
However, deviations are evident in 2014 and 2021, where actual trade only solves the problem of insufficient compatibility of information
volumes exceeded the model’s predictions—likely attributable to platforms, but also strengthens transport links, emphasizing digital
external shocks or policy changes not accounted for in the model. connectivity to promote trade growth（Ding,2018）.
Labor, efficiency reforms, and standards coordination: The
negative effect of PC3 calls for automation in logistics operations to
reduce manual errors caused by information disjointedness. In addition,
resolving trade frictions requires establishing mutually recognized
standards. Such structural reforms and training programs are critical to
overcoming operational inefficiencies and promoting integrated logis­
tics services (Kim et al., 2021).
Finally, based on the comparison between actual and predicted trade
data from 2013 to 2023, the model demonstrates strong overall pre­
dictive performance, yet discrepancies in certain years highlight the
vulnerability of China-Korea trade to external shocks such as geopolit­
ical tensions and global health crises. To address these structural and
external risks, this study proposes a three-dimensional optimization
approach. First, it recommends the construction of a "multi-hub-
decentralized" supply chain network by piloting integrated multimodal
transport models—including sea, land, air, rail, and postal service­
Fig. 5. Comparison of Actual and Predicted (2013–2022). (Unit:USD billion). s—between hubs such as Weihai and Incheon, and by strengthening the

6
Y. Zhang et al. The Asian Journal of Shipping and Logistics xxx (xxxx) xxx

Northeast Asia Trade Corridor to enhance resilience against trade Ethics statement
volatility. Second, leveraging big data and smart logistics infrastructure,
the study advocates for the development of a high-frequency digital twin The studies involving human participants were reviewed and
system for trade operations. This system would integrate real-time approved by Academic Ethics Committee. The author provided the
customs data, vessels AIS signals, and daily e-commerce transaction written informed consent to participate in this study.
data, thereby enabling predictive monitoring and early-warning capa­
bilities. Third, the establishment of institutional mechanisms such as Declaration of Competing Interest
regular China-Korea logistics coordination dialogues are proposed to
improve bilateral trade governance and reduce systemic inefficiencies. There are no conflicts of interest in this study.
These strategies not only provide technical and institutional support for All authors have read and approved this version of the article, and
China-ROK Digital Economy Partnership Agreement (under DEPA) ne­ due care has been taken to ensure the integrity of the work. Neither the
gotiations but also offer a robust, forward-looking analytical framework entire paper nor any part of its content has been published or has been
for regional economic governance, characterized by high predictive accepted elsewhere. It is not being submitted to any other journal.
accuracy and strong policy adaptability.
Data Availability
7. Conclusion
The original contributions presented in the study are included in the
This study investigates the relationship between Korea’s logistics article, further inquiries can be directed to the corresponding author.
system and its bilateral trade with China using principal component
analysis and regression modeling. Through the transformation of nine References
interrelated logistics variables into five uncorrelated components, the
model successfully identifies critical factors driving trade volume. The Bottalico, A., Vanelslander, T., & Verhoeven, P. (2022). Innovation and labor in the port
industry: A comparison between Genoa and Antwerp. Journal of Business Logistics, 43
empirical results show that logistics scale (PC1) and international
(3), 368–387.
transport capability (PC5) have significant positive effects on China- Celebi, D. (2017). The role of logistics performance in promoting trade. Maritime
Korea trade, while labor and compensation-related inefficiencies (PC3) Economics amp; Logistics, 21(3), 307–323.
Celebi, O. (2019). Logistics efficiency and trade performance: The case of China and
have a negative impact. The insignificance of Minor Cost Fluctuation
South Korea. International Trade Journal, 44(2), 45–67.
Component (PC2) indicates that short-term cost fluctuations in South Chen, H., & Hu, Z. (2016). Analysis on the link between logistics development and
Korea have been absorbed by large logistics companies through econo­ China’s Import-Export scale. Logistics Technology, 35(8), 132–137.
mies of scale, thereby minimizing trade disruptions. The insignificance Chen, W., Li, Z., & Wu, J. (2018). The effect of logistics performance index (LPI) on trade
promotion in northeast Asia. International Logistics Review, 96–112.
of Port Throughput (PC4) may reflect the saturation effect of South Duran, C., Fernandez-Campusano, C., Carrasco, R., & Carrillo, E. (2024). DMLBC:
Korean ports. Dependable machine learning for seaports using blockchain technology. Journal of
The research highlights the importance of continued investment in King Saudade University - Computer and Information Sciences, 36(1), Article 101918.
Gacuru, M., & Kabare, S. (2015). Labor capacity and organizational density in logistics
infrastructure, the development of efficient logistics networks, and re­ industry performance. Journal of Logistics and Operations, 12(3), 74–83.
forms in labor and operational management. These improvements will Guo, X., Li, J., & Huang, S. (2023). Study on trade effects of Green maritime transport
not only strengthen Korea’s trade ties with China but also enhance its efficiency: An empirical test for China based on trade decision model. Sustainability,
15(16), 12327.
overall competitiveness in regional and global markets. Jolliffe, I. T., & Cadima, J. (2016). Principal component analysis: a review and recent
Despite the model’s effectiveness, the study is constrained by its developments. Philos Trans A Math Phys Eng Sci., 374(2065), Article 20150202.
reliance on annual data and the inherent abstraction of principal com­ Kang, S. H., Nam, J. W., Sim, M. S., & Kim, Y. S. (2021). A study on the productivity and
efficiency comparative analysis of container terminal in Busan new port. Journal of
ponents. Future research can build on these findings by incorporating
Navigation and Port Research, 45(3), 138–147.
more dynamic datasets and exploring policy-oriented logistics solutions Katrakylidis, C., & Madas, M. (2019). Economic scale and trade volume: a study of South
in greater depth (Chen & Hu, 2016). Korea’s trade dynamics. Economics and Business Journal, 33(2), 101–115.
Kim, D., Cho, G., & Park, Y. (2021). Comparative Study of Policies and Systems Related
However, the current study is based on a relatively limited sample of
to Fresh Logistics in Korea, China and Japan. 무역금융보험연구 (구. 무역보험연구),
ten years, which restricts the ability to capture long-term structural 22(3), 39–51.
changes. Future research could benefit from extending the time series to Kim, H., & Lee, Y. (2016). Comparative analysis of logistics systems in korea and China:
include additional historical and recent data to improve robustness. Policy implications for bilateral trade. Korean Economic Review, 32(1), 55–70.
Kojo Ayesu, E., Sakyi, D., Arthur, E., & Osei-Fosu, A. K. (2022). The impact of trade on
Moreover, the scope of this study is primarily limited to logistics- African welfare: Does seaport efficiency channel matter? Research in Globalization, 5,
related variables. Bilateral trade is also influenced by various external Article 100098.
factors including government policy, macroeconomic conditions, cul­ Lee, S., & Song, D. (2017). Logistics service quality and customer loyalty: A Korean
perspective. International Journal of Physical Distribution Logistics Management, 47(5),
tural ties, and geopolitical developments. The incorporation of these 435–455.
broader dimensions could provide a more holistic understanding of the Liang, R., & Liu, Z. (2020). Port infrastructure connectivity, logistics performance and
China-Korea trade relationship. Future studies may expand the model by seaborne trade on economic growth: an empirical analysis on “21st-century
maritime silk road. Journal of Coastal Research, 106(SI), 319–324, 100098.
incorporating more granular data, as well as external variables including Lim, C., & Lee, S. (2021). Infrastructure and export performance: Spatial panel evidence
trade policies or digital logistics innovations (Chen & Hu, 2016). from South Korea. Regional Studies, 55(4), 601–614.
Comparative studies with other major trade partners could further Ma, Q., Lin, Y., & Zhao, L. (2020). Economic capabilities and their effect on trade flows:
A case study of China and South Korea. Trade Studies, 31(5), 105–123.
enrich the understanding of China Korea’s logistics-trade nexus.
Mahpour, A., Nazifi, A., & Amiri, Mohammadian (2021). A. Development of optimization
model to reduce unloading and loading time at berth in container ports. Iran Journal
CRediT authorship contribution statement of Science andTechnology Transaction of Civil Engineering, 45, 2831–2840.
Ministry of Trade, Industry and Energy. (2023). Trade statistics report. Retrieved from 〈htt
ps://[Link]〉.
Zhang Yue: Writing – review & editing, Writing – original draft, Shamohammadi, M., Kwon, O. K., & Lee, S. Y. (2025). Evaluation of the efficiency and
Visualization, Project administration, Methodology, Investigation, returns to scale of Korean logistics companies. Research in Transportation Business
Funding acquisition, Formal analysis, Data curation, Conceptualization. Management, 60, Article 101354.
Shen, Y., Liu, J., & Tian, W. (2022). Interaction between international trade and logistics
Nianjie Shang: Software. Liu yanting: Validation. zhang chao: Su­ carbon emissions. Energy Reports, 8, 10334–10345.
pervision. jin lu: Resources. Shikur, H. (2022). Connecting logistics efficiency and trade flows in the context of
international relations. Trade and Logistics Journal, 18(4), 78–91.
Song, D., & Lee, S. (2022). Human capital and organizational density in logistics
operations: An empirical study in South Korea. Logistics Research Journal, 10(1),
102–116.

7
Y. Zhang et al. The Asian Journal of Shipping and Logistics xxx (xxxx) xxx

Wang, F., & Liu, Q. (2021). Logistics cooperation between China and korea: A sustainable Zablotska, R., & Mammadov, K. (2017). The promoting efficient logistics services in
trade development perspective. Sustainable Development in Global Trade, 2107. international trade. Actual problems of International Relations, 131, 121–130.
Wang, Y., & Xin, L. (2020). The impact of China’s trade with economies participating in Zhang, X. (2020). Impact of inland port development on China-South korea bilateral
the belt and road initiative on the ecological total factor energy efficiency of China’s trade flow. Journal of International Trade, 62(1), 62–82.
logistics industry. Journal of Cleaner Production, 276, Article 124196.