Resilience Evaluation of Supply Chain Configuration in Export-Oriented Enterprises: Indicator Construction and Rebound Prediction Based on Transaction Data

Abstract

Under the complex and dynamic international market environment, export-oriented enterprises are increasingly exposed to external disruptions, highlighting the critical role of supply chain configuration resilience. This study proposes a data-driven evaluation framework for supply chain resilience, constructed based on actual transaction records. Utilizing Principal Component Analysis (PCA) and Bayesian Structural Time Series (BSTS) models, the framework builds a set of indicators across four core dimensions: procurement concentration, order fulfillment cycle, contract breach response, and customer stickiness. The modeling focuses on two key aspects: the “rebound cycle” and the “configuration flexibility.” Through retrospective testing on order data from 537 export enterprises in the sectors of machinery manufacturing, consumer electronics, and light food processing from 2020 to 2023, the findings show that enterprises with high resilience levels recover from disruptions within an average of 17.4 days—significantly shorter than the industry average of 25.6 days. Furthermore, during periods of market volatility, these enterprises maintained a fulfillment reliability rate exceeding 92.3%. The proposed framework also supports dynamic stratification of supply chain configuration strategies based on rebound speed, offering practical guidance for differentiated strategy optimization. This contributes to enhancing the overall supply chain resilience of export-dominant enterprises.