Dynamic Lot-sizing Decisions And Procurement Auctions Under Online Retail Auctions

The auction is an expected utility maximization problem under economic game. With the development of information technology, people begin to use the network to lead online auction. It brings new development opportunities and energies for ancient auction. With advantages of low cost, flexibility and more factors, online auctions come to more and more popular, so do the reverse auctions. In order to maximize profits, it has become an very important issue of the enterprise manage-ment how the retailer establishes the corresponding inventory and optimal lot-sizing in face of the random number of bidders. Therefore, by using the same price auc-tion mechanism, we study dynamic lot-sizing decisions under the online sequential auctions with reserve price, and inventory management under online sales auctions and procurement auctions.When having enough inventory to meet each auction, the retailer always uses sequential auctions so as to reduce negative market impact due to providing large quantities of inventory to sale. In the second chapter, we build a dynamic lot-sizing model of inventory control by using stochastic dynamic programming. It is assumed that the number of bidders in each auction is random, the customers' valuation of items are subject to uniform distribution, and the identical multi-items Vickery auction is used. When consumers' arrival process obeys Poisson distribution or Constant cases, we show that single auction revenue function satisfies a second or-der condition, the optimal inventory-scraping policy is a scrap-down-to level policy and the optimal lot-sizing policy is a monotone staircase with unit jumps policy.When the retailer observes that the inventory can't meet the demand of pe-riodic auction at the beginning of each auction, he leads procurement auction to supplement the goods. In the third chapter, we consider inventory management with two following characteristics. One is that the retailer sells products by online auctions, another is that the retailer makes procurement management by reverse auctions. For the single-period case, it is shown that the retailer's expected rev-enue is a strictly increasing and concave function of the purchased quantity under online auctions, leading to the retailer's optimal procurement strategy and the sup-pliers' optimal bidding strategy. For the multi-period case, the optimal control model is first constructed for the retailer by using stochastic dynamic program-ming. Then, the retailer's optimal procurement strategy and the suppliers' optimal bidding strategies are obtained, which are similar to the base-stock policy.