Application of text mining in identifying the factors of supply chain financing risk management

Purpose: This study aims to clarify the risk management practices of banks as supply chain finance (SCF) service providers. Design/methodology/approach: Using 4,014 evaluation and approval reports, this study constructed five risk management factors and examined their functions with secondary data. Two text-mining techniques (i.e. word sense induction, TF–IDF) were used to equip the classic routine of dictionary-based content analysis. Findings: This research successfully identified four important risk management factors: relationship-based assessment, asset monitoring, cash flow monitoring and supply chain collaboration. The default-preventing effect of these factors are different and contingent on the type of financing contexts (i.e. preshipment, postshipment). Practical implications: The empirical evidences provide practical implications for SCF service providers to manage risk. SCF service providers are suggested to pay more attention to cash flow monitoring when providing postshipment financing services and shift the focus to relationship building and supply chain collaboration when providing preshipment financing services. Originality/value: The study shows that a large volume of textual materials can provide adequate clues for researches as long as they are mined with suitable analytic techniques and approaches. Based on the results, SCF service providers can identify problems of their operations and directions for improvement. In addition, the risk management vocabulary from the E&A reports can be utilized by SCF service providers to digitize their loan approving process and, further, to facilitate the decision-makings.