Markdown Optimization via Approximate Dynamic Programming

Abstract

We consider the markdown optimization problem faced by the leading apparel retail chain. Because of substitution among products the markdown policy of one product affects the sales of other products. Therefore, markdown policies for product groups having a significant crossprice elasticity among each other should be jointly determined. Since the state space of the problem is very huge, we use Approximate Dynamic Programming. Finally, we provide insights on the behavior of how each product price affects the markdown policy.

Keywords

• Approximate Dynamic Programming
• Multinomial logit model
• Cross-price elasticity
• Markdown optimization
• Pricing