Essays in global supply chain and revenue management in the network economy.

摘要

My dissertation research includes the essays on two topics: (a) Integrated risk management (combined Operational and Financial Hedging) approaches to global production planning issues (co-work with Lingxiu Dong and Panos Kouvelis); (b) Dynamic revenue management approaches to optimizing multi-class customer demand fulfillment (co-work with Panos Kouvelis and Joseph Milner).; In the risk management essays, we study the implications of financial hedging policies on location and production planning decisions of risk averse global firms in the presence of demand and exchange rate uncertainty. We consider that a firm can choose production location between his home country and a foreign country and sell to both local and foreign markets. In the first stage a capacity, a financial hedging contract and the location of production center are decided in the presence of demand and exchange rate uncertainty. In the second stage, after the realization of demand and exchange rate, a production "allocation" decision (e.g., how many units to localize in home market and distribute to the foreign market) is made in order to optimize profits. The second stage allocation decision (referred to as an "allocation" option) is the firm's real option serving as an operational hedge of the demand and exchange rate uncertainty. We want to understand the role of the allocation option and financial hedging on the production location and planning decisions.; In the revenue management essays, we are concerned with the problem of allocating inventory over a horizon to demand from several classes of customers when partial backlogging of unfulfilled demand is possible. The customers are distinguished into several classes by the price they are to pay for the item. The probability of customers' commitment to wait is influenced by a discount the firm may offer as well as some class specific parameters.; In the first essay, demand from each customer class is modeled as a realization of a (non-stationary) random variable during each of several stages a period is divided into. The firm is able to view this demand in each stage prior to making an allocation decision on which demand to fill.; In the second revenue management essay, we focus on the deterministic demand problem with constant arriving rates on the infinite horizon problem (a generalized "EOQ" model). (Abstract shortened by UMI.)