Dr. Long Gao, A. Gary Anderson Graduate School of Management, University of California·Dynamic Supply Risk Management·2012年7月9日 星期一下午·北510 2012.07.09

[Time]:  14:30-15:30, July 9, 2012 (Today)   

[Place]: Room 510, Shunde Building   

[Topic]: Dynamic Supply Risk Management    

[Speaker]: Dr. Long Gao, A. Gary Anderson Graduate School of Management, University of California, Riverside, CA 92521long.gao@ucr.edu   

[Host]: Prof. Xiaobo Zhao


[Abstract]: We study a dynamic risk management problem in a multisupplier, multiclass customers, multiperiod inventory system. The sales are discretionary and suppliers are subject to both operational and disruption risks. The firm dynamically forecasts supply risks via advance supply signals—such as suppliers' production schedule changes, capacity commitment to their clients. We develop a Markov chain model that captures the essential empirical features of advance supply signals, including nonstationary, volatility, and dynamic evolution. We then formulate a stochastic dynamic program for integrative procurement and selling decisions that embed the Markov chain model as a short-term forecast. Through concavity, modularity and stochastic comparison we formalize four types of substitutability and complementarity critical for dynamic risk management. We show that the optimal procurement policy is characterized by supply diversication and intertemperal substitution; the optimal selling policy is driven by customer segmentation and intertemperal rationing. Moreover, two operational levers, inventory procurement and discretionary selling, are strategic substitutes; informational imperatives (of dynamic forecast) and operational imperative (of procurement and fulfillment) are strategic compliments. Our experiments on real-word data demonstrate the significant profit advantage of the optimal policy, provide operational insights on the value of signal-based forecast. 

Our study demonestrate the central role of intertemporal imperative in dynamic context. Both procurement and selling activities should be coordinated and synchronized with dynamic forecast in order to effectively mitigate supply risks. If properly implemented, our resilience driven, adaptive strategy can help eradicate many drawbacks in conventional hedging strategies and “lean” business programs. 


清华大学工业工程系
联系电话: 010-62772989
传真:010-62794399
E-mail:ieoffice@tsinghua.edu.cn
地址:北京市海淀区清华大学舜德楼5层


Copyright © 2014-2021 清华大学工业工程系 版权所有