Relating Expected Inventory Backorders to Safety Stock Investment Levels.

The objective of this research was to develop a method to define and predict the relationship between inventory performance and safety stock investment for the Defense Electronics Supply Center (DESC) inventory. DESC uses the model prescribed by DoD Instruction 4140.39 to set individual item safety stock levels. This model minimizes the sum of the variable order and holding costs subject to a constraint on the expected inventory perofmrance as measured by the number of time-weighted essentiality-weighted requisitions short (backorders). An important consideration in selecting the constraint for this model is the safety stock investment required for various levels of performance. This thesis uses multi-variate regression analysis and forecasting techniques to predict the relationship between expected performance and required investment. The author concludes that this method produces accurate predictions of the relationship. The recommended model produced average absolute errors of about three percent when tested against historical data from the DESC inventory. (Author)