Abstract: In order to study the inventory optimization problem of dual-channel supply chain, a dual-channel supply chain system of defective goods consisting of an online retailer and a traditional retailer is considered. It allows different retailers to rebalance the inventory level by lateral transshipment strategy, introducing customers switching between channels, and using Markov chain to depict the retailer's inventory status. Finally, a minimum inventory cost model is established, which is solved by genetic algorithm. By comparing the two schemes of allowable transshipment and non-allowable transshipment, it is concluded that in the dual-channel supply chain of defective products, the inventory and total cost can be reduced by adopting transshipment strategy between two channels; and that with the increase of customers switching probability and product defect rate, the system performance induced by the transshipment strategy is decreased.

Key words: customers switching, Markov chain, dual-channel, transshipment, defective products