Integrating simulation and decision trees through blockchain-enabled data sharing to prevent the cash flow bullwhip effect in supply chains

Working capital mismanagement poses significant challenges to supply chain (SC) operations, threatening the stability and viability of businesses worldwide. One manifestation of this issue is the cash flow bullwhip (CFB) effect, which refers to the amplification of working capital variability relative to demand variability as it propagates upstream in the SC. Blockchain-enabled data sharing and decision trees trained on data generated by discrete-event simulation are potential yet unexplored solutions to address the CFB effect. This study fills this gap by investigating the effectiveness of blockchain-enabled data sharing and the integration of discrete-event simulation with decision trees in mitigating the CFB effect. The analysis focuses on a three-echelon manufacturing-retail SC. However, the findings are applicable to other SC types that experience the CFB effect. Blockchain provides visibility into end-customer demand and working capital policies across SC tiers by enabling data sharing. The shared data serve as inputs into a discrete-event simulation model that generates dynamic scenarios to train decision trees. Findings demonstrate that demand forecasting based on end customers’ needs, facilitated by blockchain, significantly reduces the CFB effect. Additionally, combining this forecasting with uniformly applied, increasing cash collection policies across all SC members, also coordinated by blockchain, can prevent the CFB effect. Decision trees provide interpretable and actionable rules for setting working capital policies, highlighting the importance of regulating inventory policies at the middle echelon of the SC to prevent the CFB effect. This study offers managerial recommendations to address the CFB effect in SCs.