Safety Stock

Abstract

Safety stock is the extra inventory to hold for an item for protection against demands exceeding the forecast. This type of stock concerns entities where the demands of the future are not known until they happen, like in distribution centers, stores and dealers. In plants, where the production schedules are set in advance, safety stock is usually not needed. The typical way to measure the variability in the forecasts is by the standard deviation of the one month ahead forecast error. A relative measure is the coefficient of variation. Two common methods of generating the safety stock are: the service level (probability not out of stock), and the percent fill (ratio of demand filled over total demand) methods. Both methods are also sometimes referred as the service level method. The normal distribution is used primarily to generate how much safety stock to have available. This chapter shows how the truncated normal distribution can also serve this function. The truncated normal has many shapes and includes only portions of the right-hand-side of the standard normal.

Summary

Safety stock is one of the most important measures in inventory control . The two common ways to measure the safety stock are the service level method and the percent fill method . Two probability distributions are described on how to determine the amount of safety stock needed, the standard normal distribution and the truncated normal distribution . No doubt, the standard normal is far more known and used. Although the truncated normal distribution is not known or used by most practitioners, it is the more appropriate distribution to apply when seeking the safety stock for each sku in the inventory. To use the normal distribution, when the truncated normal is appropriate, will result in a lower level of service (service level, percent fill), than is planned.