Reorder Point Calculator

A reorder point is the trigger level in a continuous review policy: whenever on-hand plus on-order stock falls to the reorder point, a replenishment of fixed size is released. Only the exposure window matters, which is the lead time, because that is the interval during which no further goods can arrive. The point therefore has two parts - expected demand over the lead time, d multiplied by LT, and a buffer sized to cover the variability of that demand. The statistical buffer combines two independent sources of noise under a square root: lead time multiplied by the variance of daily demand, plus squared average demand multiplied by the variance of lead time. Scaling by a z value converts it to a chosen cycle service level, with z equal to 1.28 for 90 percent, 1.645 for 95 percent and 2.326 for 99 percent. The second term is usually the one that dominates. When a supplier lead time swings by a day or two around a mean of a week, the d squared multiplied by lead time variance component typically outweighs the demand term entirely, which is why supplier reliability generally buys more than demand forecasting accuracy. The reorder point answers when to order; EOQ answers how much. Assumptions to watch are that demand is roughly normal and serially uncorrelated, that lead time and demand are independent, and that inventory position rather than just on-hand stock is used for the comparison - failing that last point causes duplicate orders during long lead times.

Trigger-level replenishment predates any formal statistics. The two-bin system, in which stock is split into a working bin and a reserve bin and reordering begins when the working bin empties, was standard shop-floor practice in early twentieth-century factories and encoded exactly the same logic in physical form. Formal treatment arrived with the wartime and post-war operations research movement. Kenneth Arrow, Theodore Harris and Jacob Marschak published Optimal Inventory Policy in Econometrica in 1951, framing replenishment as a stochastic optimisation problem, and Thomson Whitin The Theory of Inventory Management appeared in 1953. Herbert Scarf proved in 1960 that the (s, S) policy - order up to S whenever inventory falls to s - is optimal under fairly general cost assumptions, which gave the reorder point a rigorous foundation. Robert Goodell Brown, working at Arthur D. Little, translated this into usable procedure in Statistical Forecasting for Inventory Control in 1959 and later volumes, popularising the practice of driving safety stock from a forecast error standard deviation and a service-level factor. Through the 1960s these rules were programmed into early inventory systems, and Joseph Orlicky Material Requirements Planning in 1975 argued that dependent demand components should instead be timed from a bill of materials rather than reordered from a statistical trigger. Contemporary ERP systems keep both: MRP logic for dependent items and reorder point logic for independent-demand and low-value parts.