Abstract

This study investigates how container (BPC) and non-container (BPB) cargo throughput at Busan Port responds to structural shocks in global and domestic macroeconomic variables, and how their relative importance changes over time. The selection of variables was based on a combination of a Systematic Literature Review (SLR) and semi-structured interviews (SSI). The econometric analysis applied a stepwise framework, employing cumulative Impulse Response Functions (IRF) and Forecast Error Variance Decomposition (FEVD) based on Vector Error Correction Model (VECM) estimations, as well as Historical Decomposition (HD) for specific crisis periods. While IRF and FEVD were used to capture dynamic properties under general conditions, HD was employed to identify and compare the structural contributions of variables during particular crises, namely the global financial crisis (2008Q1–2010Q4) and the COVID-19 pandemic (2020Q1–2022Q4). The dataset consisted of quarterly observations from 2005 to 2022, and included BPC and BPB as dependent variables, alongside global factors (Baltic Dry Index, BDI; China Containerised Freight Index, CCFI; World Export and Import Indices, MEI/MII; China’s export and import volumes, CEV/CIV) and domestic factors (Korea Composite Index, KCI; Korea’s export and import values, KEV/KIV; Korea’s GDP, KGDP; and the KRW/USD exchange rate, KER).The key findings are as follows. First, IRF results showed that BPC exhibited moderate positive cumulative responses to CIV and MEI/MII, while KER produced consistently significant negative responses across all lags, suggesting that depreciation of the won structurally constrained container throughput. Domestic economic variables (KGDP, KEV, KIV, KCI) displayed largely flat or insignificant responses. By contrast, BPB showed consistent negative cumulative responses to MII, positive responses to MEI and CIV, and modest positive responses to KER, which were statistically significant at certain lags.Second, the FEVD numerically confirmed these differences in sensitivity. For BPC, the explanatory power of its own variable decreased from around 40% initially to about 29% in the long term, with the influence of exogenous variables expanding. Among these, KER accounted for the largest long-term contribution (approximately 35%), while MEI, MII, and CIV each sustained meaningful explanatory power of about 7–8%. In contrast, most domestic variables contributed less than 5%. For BPB, the explanatory power of its own variable declined markedly from around 12% to 3% in the long term. MII accounted for the highest contribution (around 30%), followed by CIV (13%), MEI (11%), with BDI and CCFI also providing notable explanatory shares.Finally, HD analysis was conducted to identify structural contributions during specific crisis periods. During the global financial crisis (2008Q1–2010Q4), the decline in BPC was mainly driven by negative contributions from KER and MII, while recovery was supported by increasing positive contributions from MEI and CIV. For BPB, negative contributions from freight indices (BDI and CCFI) were dominant during the downturn, whereas recovery was driven by positive contributions from BDI, CIV, and MII, with KER acting as a buffer in certain phases. During the COVID-19 pandemic (2020Q1–2022Q4), BPC was constrained primarily by the persistent negative contributions of MII. KER initially provided a modest positive buffering effect, but later shifted to a negative influence, while BDI played a mitigating positive role after mid-2021. For BPB, MEI exerted a strong negative shock in 2020, but recovery in 2021 was supported by the positive contributions of MII, with BDI and CCFI also playing favourable roles. CIV, however, showed differentiated patterns, shifting to negative contributions in certain quarters.In sum, Busan Port throughput is structurally sensitive to global variables, though the transmission channels and relative importance differ by cargo type. The financial crisis was characterised by the dominance of financial and monetary channels, such as exchange rates and freight indices, while the pandemic was driven more by real economy channels, including demand contraction and supply chain bottlenecks. Methodologically, this study contributes by presenting a “crisis-sensitive dynamic analysis framework” (SLR–SSI–VAR/VECM–IRF/FEVD/HD), enabling the integrated assessment of response structures under general conditions (IRF and FEVD) and structural contributions during crises (HD). Practically, the study proposes the establishment of an early-warning and response system using MII, MEI, CCFI, BDI, CIV, and KER as key monitoring indicators, differentiated operational strategies for container and non-container cargo, and crisis-specific policy priorities—exchange rate stabilisation and export recovery for financial crises, or global demand monitoring and supply chain resilience for real economy crises. Limitations remain, particularly the focus on a single port and the use of quarterly data. Future research should therefore consider multi-port comparisons, the inclusion of non-quantitative factors, and a wider range of global events to enhance external validity and policy applicability.

Awarding Institution(s)

University of Plymouth

Supervisor

Saeyeon Roh, Minchul Sohn

Document Type

Thesis

Publication Date

2026

Embargo Period

2026-01-29

Deposit Date

January 2026

Download

Share

COinS