Taming the Long Tail: The Gambler's Fallacy in Intermittent Demand Management.

This article focuses on optimizing inventory management for long tail products with intermittent demand by accounting for the finite horizon effect and the gambler's fallacy phenomenon. It demonstrates that traditional inventory models, which assume infinite horizons and unbiased demand distributions, are inadequate for such products due to biased demand estimates arising from limited observation periods. The authors propose a staggered base stock (SBS) policy that strategically delays replenishment timing and adjusts order quantities based on biased joint distributions of demand size and interarrival times, improving key performance metrics such as fill rate, cost per cycle, and EBITDA margin. Empirical validation using real intermittent demand data from a heavy machinery parts distributor shows that the SBS policy outperforms established benchmarks like Croston's method and the ITE policy, achieving lower costs and inventory levels. The study highlights managerial implications including the benefits of embracing biased demand estimates and replenishment delays in finite horizon settings, and suggests future research directions involving continuous-time models, nonstationary demand, and multi-item inventory systems.