Markdown Pricing

Markdown Pricing is the operations research and retail-pricing practice of setting, over the selling horizon of a seasonal or perishable item (fashion collection, fresh food, electronics with limited shelf life), a typically monotone-decreasing price trajectory. The decision balances margin (a high price gives high per-unit profit but slow sell-through) against clearance risk (unsold stock at season-end becomes dead stock — salvage, donation or disposal). The foundational empirical reference is Smith and Achabal (1998), who showed in large US retail chains a 5-15% gross-margin uplift over fixed-rule markdowns; Bitran and Mondschein (1997) built the canonical stochastic dynamic-program formulation with Poisson demand; Elmaghraby and Keskinocak (2003) reviewed the field; Caro and Gallien (2012) deployed the approach in a real fast-fashion pilot reporting 5-8% clearance-margin uplift and 30% dead-stock reduction. Methods include deterministic dynamic programming, stochastic DP, multi-product MIP for cannibalisation, and Bayesian learning + Thompson sampling for daily online repricing. Constraints typically include monotone-decrease (price does not bounce), a maximum-discount cap for brand-image protection, and a minimum-price floor. Markdown Pricing is distinct from Revenue Management (fixed capacity, up-down pricing) and Newsvendor (single-period order quantity before the season).